The common but differentiated responsibilities of states to assist and receive ‘climate refugees’

Introduction

Since the signing of the United Nations Framework Convention on Climate Change (UNFCCC) in 1992, debates about state burden sharing have been preoccupied with the allocation of responsibilities for mitigation, adaptation and climate finance. This paper draws on these debates to address the relatively neglected question of the responsibilities of states to assist and receive international migrants who are forced to leave their territory as a result of the unavoidable impacts of climate change. This is a politically opportune time to be asking such questions given that the parties to the UNFCCC agreed in 2013 to establish the Warsaw International Mechanism for Loss and Damage associated with Climate Change Impacts (UNFCCC, 2013). This mechanism is expected to serve as a vehicle to channel financial, technical and other support to developing countries that are particularly vulnerable to the impacts of unavoidable climate change (including slow and sudden onset events), which extends to population displacement, climate-induced migration and loss of statehood. The structure and mandate of the Loss and Damage mechanism will be reviewed in 2016, which provides an opportunity to ask whether the agenda should be expanded to include what might be called ‘the responsibility to receive’ on the part of host states, and the rights of climate migrants to choose their host when internal migration is no longer an option. ¹

Migration induced by climate change, like other forms of migration, sits along a continuum ranging from those seeking a better life to those who are moving to save their lives. While the Warsaw mechanism should extend to cover all cases of loss and damage associated with unavoidable climate change impacts, the particular focus of this discussion will be confined to international climate migrants who are forced to leave their territory permanently due to rising seas and other unavoidable climatic impacts. ² These climate migrants, most of which are expected to come from small island developing states (SIDS), do not qualify as refugees under the 1951 Refugee Convention because they are not fleeing political persecution. Moreover, some SID leaders, such as Kiribati’s President Anote Tong, have rejected the refugee label as undignified, since island people do not wish to move, and they do not wish to be treated like political refugees (ABC, 2014). Nonetheless, I use the term ‘climate refugees’ to describe this particular group in order to focus on the similarities and differences with political refugees that might be ethically and politically relevant in determining state responsibilities and refugee rights (without endorsing this label for the proposed protection regime).

The UNFCCC’s burden sharing principles of ‘equity and common but differentiated responsibilities and respective capabilities’ (hereafter CBDR-RC) apply to both mitigation and adaptation (see Articles 3(1), 4(1) and 4(3)). On one view, international migration may be regarded as the ultimate form of adaptation to devastating climatic impacts and therefore the UNFCCC principles should extend to guiding the allocation of the burden sharing for climate refugees (e.g. Bierman and Boas, 2008). ³ Yet it can also be argued that this particular form of migration is a response to the limits of in-country adaptation and, in any event, is a form of loss and damage that cannot be reduced to adaptation. The preamble to the decision establishing the Warsaw Mechanism acknowledges that loss and damage ‘includes, and in some cases involves more than, that which can be reduced by adaptation’ (UNFCCC, 2013: 6). In the negotiations leading to the establishment of the Warsaw mechanism, China and the G77 had argued that loss and damage should form a third pillar of the Convention, in addition to mitigation and adaptation, while the US insisted that it should be considered only as part of the adaptation agenda (ENB, 2013: 18; Zhu, 2013). The US prevailed, with the final text declaring that the mechanism would fall ‘under the Cancun Adaptation Framework’ but the quid pro quo was that the US agreed to the 2016 review (UNFCCC, 2013). The words ‘liability’ in relation to loss and damage are conspicuously absent from the text and the questions of who is responsible for providing assistance and on what basis have not been determined.

This uncertainty calls for critical reflection and guidance on the nature of state responsibility for climate refugees. In particular, should state responsibility be understood as a simple humanitarian responsibility to provide support and a safe haven, which is left to the discretion of each state (like the current arrangement for accepting a particular quota of political refugees)? Or should state responsibility be understood as a stronger legal and/or moral responsibility to provide redress for loss and damage for which the host state is causally responsible? Should state responsibilities to provide financial and other support to climate refugees, and to host them, be connected or can they be separated? And how might various permutations of these responsibilities affect a possible right on the part of refugees to choose their host?

There is now a well-developed debate among climate ethicists and climate justice theorists on the question of how to assign responsibility to states for mitigation and adaptation (e.g. Caney, 2005, 2010; Gardiner, 2011; Meyer and Roser, 2010; Moellendorf, 2014; Page, 2012; Vanderheiden, 2008, 2011). Most of this debate has taken place within the contours of liberal moral theory (broadly conceived), although there are also some distinctive strains of critical political ecology with Marxist roots (e.g. Parks and Roberts, 2010; Roberts and Parks, 2007). There are also well-worn lines of debate among the parties to the UNFCCC over the interpretation and application of the convention’s principles of CBDR-RC. The problem is that while the debate among climate ethicists and climate justice theorists has become increasingly refined over the past two decades of climate negotiations, the division among many of the key players (particularly between the Umbrella group led by the US and the so-called BASIC group of China, India, South Africa and Brazil) has widened and hardened. This yawning gap between ‘the ideal and the real’ provides a serious challenge for those seeking justice for climate refugees, who are among the least responsible for, and the most vulnerable to, the impacts of climate change. Although many of the scholarly debates are informed by the real-world climate negotiations, there have been very few attempts by moral philosophers and normative political theorists to bring the ethical and real-world political debates into any kind of close, critical encounter for the purposes of developing action-guiding norms that have some reasonable prospect of implementation (an exception is Pickering and Barry, 2012).

The article seeks to arrange such an encounter in order to expose the yawning chasm between the most mooted ideal and real-world arguments relating to state responsibilities, identify the ethical and political trade-offs associated with these different framings of responsibility and negotiate a set of action-guiding norms that is likely to maximise the protection for climate refugees. The normative method draws inspiration from the tradition of Critical Theory, which can be traced back to Marx, which seeks to understand rather than bracket the course of history in order to find emancipatory possibilities that are immanent in real-world developments. Andrew Linklater has called this method ‘praxeology’, which entails critically reflecting upon and harnessing those moral resources within existing social arrangements that might enable new forms of community with higher states of freedom (Linklater, 1998: 5), or in this case, forms of community that do not entail the obliteration of freedom for some. Praxeology requires making practical and context-dependent judgments about which norms are most likely to provide the best protection for climate refugees given feasibility constraints. This includes probability assessments that require an understanding of the history of the international climate negotiations and the relationships between the parties whose agreement is required to implement reforms. For reforms to be feasible they must be capable of being implemented politically, which means that there must be historical agents with the capability and the will to carry out the reform, and that the reform must be relatively stable rather than short-lived (see, for example, Gilabert and Lawford-Smith, 2012: 813). Acknowledging feasibility constraints does not mean limiting ethical horizons but rather using ethical horizons as a guide to finding more creative and practical ways of addressing some of the motivational constraints that stand in the way of better protection for climate refugees. In this case, the relevant agents are the parties to the UNFCCC. However, other agents (such as domestic actors or international organisations) are also relevant insofar as they have the capacity to influence states.

Finally, judgments that seek to reconcile what is ethically desirable with what is politically feasible are necessarily time sensitive insofar as political opportunities and ‘policy windows’ can open and close in ways that can alter strategic assessments of political feasibility. In this case, while major flows of climate refugees are not imminent they are nonetheless inevitable, and the particular policy window that is identified is relatively short. The analysis that follows focuses on which action-guiding norms might gain political traction by the time the parties to the UNFCCC embark upon the scheduled review of the Warsaw Mechanism on Loss and Damage in 2016. Until then, the negotiations are expected to remain focused on the provision of financial and other support. By 2016 the parties may be ready to revisit and expand the agenda to consider and clarify the question of the responsibility to receive.

The primary focus of the encounter will be the most mooted ethical principles for fleshing out the meaning of CBDR-RC, namely, the polluter pays principle (PPP), the beneficiary pays principle (BPP) and the ability to pay principle (APP). These principles will be situated in the context of real-world debates about CBDR-RC in the international climate negotiations and, where relevant, the real-world politics of managing flows of political refugees. I conclude by defending a normative package that separates the responsibility of states to provide financial and other assistance to climate refugees from the responsibility of states to receive them.

Historical responsibility

The simplest ethical principle for causal responsibility for climate change is the PPP, according to which the polluter should pay for the costs of pollution. This principle was first formulated by the OECD in 1972 as an economic principle to ensure the ‘internalisation’ of negative environmental externalities in the price of goods to avoid distortions in international trade and investment, and it has since been interpreted to mean that the polluter is liable for the costs of pollution that have been externalised onto third parties (Attapattu, 2006: 440, 448–451; OECD, 1972). In the case of climate refugees, it would apply as an ex post obligation by states to compensate climate refugees on the basis of each state’s relative causal contribution to the loss and damage suffered, measured in terms of total cumulative emissions rather than current aggregate or per capita emissions.

However, the PPP found very little favour with the negotiators of the UNFCCC – indeed, the Alliance of Small Island States was the lone supporter (Ashe et al., 1999: 215). The main resistance by developed countries arose from concern over its legal implications, including a lack of consensus over whether the PPP imposed an ex ante obligation to internalise the environmental costs of production or an ex post obligation to compensate victims on the basis of strict liability (Attapattu, 2006: 454–455; Bodansky, 1993: 501). China and the G77 rejected the PPP because it was ‘development blind’, preferring to focus instead on the historical responsibility of the ‘early developers’ and the development needs of the ‘late developers’.

Many climate ethicists have also questioned the fairness of the PPP to ground historical responsibility as a form of strict liability. The most common objection is that of excusable ignorance. Why should the earlier industrialisers be made responsible for the cumulative emissions that they have generated since their industrialisation when they did not know of their harmful effects? Critics have also pointed to the difficulties of measuring relative contribution, the fact that past carbon polluters are now dead, and that the PPP cannot address climate-related harms that cannot be attributed to particular polluters (e.g. Caney, 2010; Gardiner, 2011; Meyer and Roser, 2010; Page, 2012; Vanderheiden, 2008, 2011).

In response to the main objection of excusable ignorance, some climate ethicists have suggested a less demanding application of the PPP to the effect that developed countries should be held responsible for their historical emissions from the date at which the problem was known (e.g. Gardiner, 2011). This is not as straightforward as it might first appear, since knowledge of climate change and its impacts has evolved slowly and fitfully over more than a century. If we take the most ‘lenient’ version of this approach and argue that, at the very least, states have known about the problem of climate change since they started negotiations for the UNFCCC in 1990, then this would result in the US and China taking on the largest burden to support and to receive climate refugees since together they account for the largest share of cumulative emissions since 1990. Given the rapid growth in China’s aggregate emissions it would soon overtake the US in bearing the largest burden. We would expect strong resistance from China, given its unmet development needs and the fact that it is much more vulnerable to climate change than US.

In the international climate negotiations, the ‘development sensitive’ case for historical responsibility for mitigation and adaptation waged by China and the G77 has been based on the right to develop and the notion of a ‘carbon debt’ or ‘ecological debt’ owed by the North to the South for using up a disproportionate amount of ‘ecological space’ or greenhouse gas absorption space in the Earth’s atmosphere. Many developing countries, led by Brazil and strongly supported by India and China, have long maintained that responsibility for mitigation according to CBDR-RC should be understood in terms of a state’s historical (i.e. cumulative) emissions, unswayed by arguments about excusable ignorance. Although China’s aggregate emissions are now larger than the US and its cumulative emissions are growing, the US has made a much greater cumulative contribution, and it is the growing stock of greenhouse gases in the atmosphere rather than the size of annual aggregate emissions in any one year that ultimately will determine the degree of planetary warming. From the standpoint of developing countries, this is why developed countries are under an obligation to take the lead in mitigation under Article 3(1) of the UNFCCC, and it likewise underpins the obligation of developed countries to contribute to climate finance to pay developing countries to decarbonise and adapt to climate change under Articles 4(1) and 4(4). The alternative would be to ‘kick the ladder down’ and deny the developing world their legitimate right to develop.

These arguments have also found support among critical political ecologists, who have drawn on Marxist dependency theory and World System theory to frame the problem of the carbon debt as but one manifestation of a broader ecological debt resulting from the ‘ecologically unequal exchange’ in the flows of energy, materials, resources and emissions between North and South. Ecologically unequal exchange is understood as the legacy of the colonial period and, more recently, the shifting international division of labour between countries in different stages of development in an increasingly integrated global economy (Andersson and Lindroth, 2001; Parks and Roberts, 2010; Rice, 2007; Roberts and Parks, 2007). On this analysis, the notional carbon debt arises not simply from the longer period of industrialisation enjoyed by developed countries but also from the ‘offshoring’ of emissions made possible by increasing international specialisation and exchange, whereby emissions-intensive industries are increasingly relocating to ‘Factory Asia’ and other rapidly growing economies of the developing world. This is camouflaged by the UNFCCC’s production-based accounting system, which only measures the emissions that are produced within the territory of states without adjusting for trade. This has enabled an ‘outsourcing’ of both emissions and responsibility; that is, as many developed countries move towards a low-emission service economy their consumers are able to enjoy the benefits of imported emissions-intensive goods without having to take responsibility for the emissions, which are attributed to the developing countries in which they are produced. As we shall see, the structural character of this problem makes it increasingly difficult to attribute responsibility to particular agents.

However, as Pickering and Barry (2012) have argued, while it is possible to develop a defensible ethical account of historical responsibility it is clear from the history of the climate negotiations that this argument would not provide a politically fruitful basis for advancing the negotiations on a new climate agreement or climate finance contributions. The same can be said for negotiations for a new agreement or mechanism to manage climate refugees. As Pickering and Barry further note, ‘many developing countries (including highly vulnerable small island states) have themselves eschewed the rhetoric of climate debt in favour of other strategies for boosting collective responsibilities to address climate change’ (2012: 679). Whether this is out of political necessity and/or a growing split within the G77 between the major emerging emitters and the Least Developed Countries is not clear but is most likely both reasons, although it does not necessarily follow that the same reasoning would apply to loss and damage. What is clear is that there is a growing sense of frustration and despair on the part of the most vulnerable states, including SIDs, with the intransigence of all the major emitters.

Beyond the liability model: Relative capability and the APP

The political difficulties and measurement complications associated with mounting any kind of argument for historical responsibility suggest that it would be more productive to abandon the fault-based or unjust enrichment approach to allocating the ‘responsibility to receive’ climate refugees, since it clearly fails the feasibility test. This leads directly to the APP, which would require all states to do what they can within the limits of their respective ability to assist climate refugees, irrespective of their causal contribution to the problem. Yet the greatest virtue of the APP is also the greatest drawback from the standpoint of climate refugees. On the one hand, it is much less complicated and contentious than the PPP or BPP to frame state responsibility for climate refugees as a collective responsibility of all states that is differentiated in accordance with relative capabilities. On the other hand, this is only likely to be politically accepted if relative capability is determined subjectively by each state rather than according to objective measures such as relative GDP or the human development index. This would effectively convert state responsibility into a charitable responsibility to provide humanitarian relief, leaving it to states to make their own subjective judgments as to how many climate refugees they are able to receive, and how much they are able to spend on providing relief, in the context of the wide range of competing demands on their budgets and resources. Given the discretionary nature of this judgment, then even the richest states may decide that providing relief to climate refugees must be a low priority in the light of their competing responsibilities. It is precisely because of the discretion attached to humanitarianism that climate justice advocates prefer the liability model, since agents who are found morally or legally liable to pay damages to a victim as compensation or restitution for harm suffered, and who are able to pay, cannot appeal to cost or competing considerations to absolve them of their liability.

We have seen that a simple liability model grounded in the PPP or BPP faces major challenges in terms of political acceptance and calculability. The situation becomes further complicated when undeserved harm is systematically produced by social structures rather than, or in addition to, the wilful or reckless actions of individual agents. There is now a growing body of critical social and political theory that has called into question the usefulness of a simple ‘liability’ model of responsibility that looks for a clear chain of causation between guilty culprits and innocent victims for particular harms according to the liberal moral grammar of responsibility – especially in a global risk society (Beck, 1995; Lavin, 2008; Young, 2011). In situations of complex interdependence, this is seen as a hopeless task because it deflects attention from the main ‘culprit’, which are the social structures that enable the systematic generation of harm. Most notably, these include capitalism but also the state system, both of which create incentives for firms and states to privatise gains and externalise costs onto third parties in space and time. For Ulrich Beck, conventional models of individual responsibility serve to normalise the system of ‘organised irresponsibility’ that generates ecological risks (Beck, 1995: 64–65). For Iris Marion Young, ‘the liability model of responsibility’ naturalises unjust social structures and therefore deflects attentions from the collective action that is required to transform them; it also singles out guilty agents, and thereby absolves everyone else from taking responsibility; it produces highly defensive and politically unproductive behaviour on the part of those who are singled out; and it is backward looking and deflects attention from the future political work that needs to be done (Young, 2011: 105–106).

Young’s account of political responsibility to transform unjust social structures is grounded in a ‘social connection model’ which arises ‘not from living under a common constitution, but rather from participating in the diverse institutional processes that produce structural injustice’ (Young, 2011: 105). However, she recognised that not all agents are equally capable of exercising political responsibility and she provides some broad parameters for thinking about how it might be differentiated according to each agent’s relative power, privilege and capabilities (Young, 2011: 142–151), which are largely (although not exclusively) derived from their social location within the social structures that need to be transformed. For present purposes, we may collapse ‘power, privilege and capabilities’ into a general notion of relative ability to transform social structures so that they no longer generate undeserved harm.

Although Young’s primary focus was sweatshop labour in developing countries rather than climate change, and the political responsibility of citizens rather than states, her arguments provide a fresh perspective on the APP by directing attention to the structural character of globally generated risks such as climate change. However, this fresh perspective is sobering because it also highlights the problem that those social agents (such as states, corporations and the owners of capital generally) with the greatest capability to transform structures typically (although not invariably) have the weakest motivation to do so because of the benefits they enjoy under such structures by virtue of their privileged social location. Conversely, those who benefit least or suffer the most from the so-called side effects of production and consumption, and inter-state rivalry, typically have the least capacity to transform social structures or otherwise alleviate their plight by virtue of their more marginal location with such structures. This is certainly the case for climate refugees and herein lies one of the greatest tragedies of the climate challenge. This helps to explain why Young chose to focus on the political responsibility of citizens rather than states; she considered the latter to be insufficiently motivated to address structural injustice in the absence of the political mobilisation by citizens (Young, 2011: 151).

In all, Young’s analysis highlights complications and challenges facing climate refugee advocates by highlighting the inverse relationship between capability and political motivation. Moreover, in focussing on the political responsibility to transform social structures she devoted very little attention to the question of responsibility for loss and damage for structural injustices that are already locked in. That said, a more critical reading of Young helps us to see a major irony in her critique of the liability model and the potential future importance of the liability model. As Anthony Langlois (2014: 56) has pointed out, if we examine more closely what has made the anti-sweatshop campaign a success in improving the situation of workers in the garment industry, it was not the ‘forward-looking’ responsibility taken by transnational corporations to improve their human rights practices for the future, since the codes of practices that were developed were weak and primarily motivated by a concern to protect brand reputation. Rather, it was the application of the liability model on behalf of workers, which required backward-looking assessments of fault and the determination of compensation.

Finally, Young’s account differs from a straightforward liberal cosmopolitan approach towards those who suffer undeserved harm, which depends on nothing more than our common humanity as a good enough reason for everyone to render as much assistance as possible within the limits of our capability. For Young, political responsibility still depends on finding a nexus of some kind (in this case a ‘social connection’) between capable agents and victims of undeserved harm, albeit mediated by social structures. This would mean that highly capable agents who disconnect from such social structures would presumably have no political responsibility whatsoever towards those suffering undeserved harm.

However, those who have grappled specifically with the structural character of the undeserved harms generated by climate change have defended a cosmopolitan interpretation of the APP. For example, Darrell Moellendorf, like Young, rejects a retrospective account of remedial responsibility based on the PPP or BPP in favour of a prospective account of social responsibility based on the APP (2014: 157). He argues it is enough merely to be a privileged member of a certain kind of arrangement or institutional order, rather than the recipient of a benefit produced by an historic injustice, to ground ‘social responsibility’ for mitigation (Moellendorf, 2014: 173–174). Whereas Young’s account has its roots in Marxist political economy, Moellendorf draws on the Rawlsian idea that the fairness of a cooperative arrangement should be judged by its effects on the least well off. This means that the allocation of the burden of mitigation to states should not set back the prospects of the poor living in the least-developed and developing countries. Accordingly, mitigation responsibilities should be assigned to states on the basis of the APP, measured in terms of a state’s level of human development (Moellendorf, 2014: 176).

In a similar vein, Bukovansky et al. (2012) have developed an ethic of ‘special responsibilities’ for climate change that links structural vulnerability, dependency and capacity in ways that are specific to particular problems. This is a distinctive type of cosmopolitan ethic that seeks to respond to a world characterised by the formal equality of states and their inequality of material capability. This is a forward-looking approach insofar as it focuses on those states (and other agents) with the greatest relevant capability to alleviate the plight of the most vulnerable simply and precisely because the most vulnerable are largely at the mercy of their actions or inactions (Bukovansky et al., 2012: chapter 6). In relation to mitigation these are the greatest aggregate emitters (essentially, the top 20 emitters, who are collectively responsible for around 80% of global emission), since they have the greatest relative capability to address the problem. However, the relevant capability in relation to the loss and damage suffered by climate refugees would be the relative capability to provide financial, technical and other assistance to resettle and this would be a much larger group of states than the top 20 emitters and could extend to all countries, adjusted for level of development (and exempting the least developed countries).

Focussing on those states with the financial and technical capability to provide assistance and relief to climate refugees has the merit of focussing on what is required to get the job done for climate refugees. But one can anticipate objections from the so-called clean wealthy countries to the effect that the ‘dirty wealthy countries’ should be required to pay more. Bukovansky et al. accept these qualifications to their ‘relevant capability’ approach by suggesting that special responsibilities should be weaker in relation to the most capable states that have neither caused, nor benefited from, the vulnerability of others, and stronger in relation to the most capable states that have made the greatest causal contribution to, or enjoyed the greatest benefits arising from, the relevant vulnerability (Bukovansky et al., 2012: 219–220). Of course, these concessions reintroduce all of the problems we have noted with the PPP and BPP, albeit through the back door. Moreover, this cosmopolitan ethic of special responsibilities does not address the difficult question of why the most capable states would be motivated unilaterally, or by multilateral agreement, to accept this kind of responsibility towards the most vulnerable as a moral or legal duty as distinct from a voluntary act of charity.

Indeed, since the negotiations launched at the 17th conference of the parties (COP17) in Durban in 2011 for a new climate agreement to be signed in Paris in 2015, to be ‘applicable to all’, the parties have effectively given up on trying to reach a principled agreement on the meaning and allocation of CBDR-RC. Instead, the negotiators have found a way of stepping around the impasse over historical responsibility by leaving it to each party to decide for themselves. That is, instead of negotiating ‘internationally legal binding commitments’ on the basis of a principled interpretation of differentiated responsibilities and capabilities with respect to mitigation and adaptation, the parties have agreed to determine their own ‘intended nationally determined contributions’ (INDCs) for the post-2020 period according to their own national circumstances and their own understanding of what they consider to be their fair share. This has meant that the political and ethical debate will shift towards the design of an ongoing review process for these INDCs, including what matters might be included in national contributions, how much flexibility should be accorded states in determining these matters, how these are to be judged and what kind of cycle of review might generate continuous improvement in contributions. This leaves the door open for parties to include as part of their INDCs a financial contribution to the Warsaw International Mechanism to assist the resettlement of climate refugees. This might be extended to include hosting responsibilities, depending on the choices made by climate refugees; indeed, Risse (2009: 294) has suggested that one country could take a whole nation and treat it as part of its contribution to adaptation. The adequacy of these and other related commitments could be included in a technical and equity review process.

Given the odium surrounding the PPP and the BPP and the state of play in the climate negotiations, including the modest progress thus far on the Warsaw International Mechanism on Loss and Damage that demonstrates at least an implicit promise by the parties to provide assistance to climate refugees, a capability-based approach to the state responsibility to assist, based on the APP, appears to be the only politically feasible option in the near term. However, the APP would also appear to rule out any general right on the part of climate refugees to choose their host state since not all states would have the same capability to receive and some states would lack such capability. Moreover, the sobering history and politics of the Refugee Convention would, in any event, suggest that states would have little motivation to grant such a right to climate refugees.

To develop an ethically desirable and politically feasible account of the right of refugees to choose their host state it is necessary, first, to distinguish climate refugees from political refugees and, second, separate the two dimensions of state responsibility (to assist and to receive).

The right to choose

The plight of climate refugees is analogous to political refugees in one major respect: they are moving to save their lives, which includes avoiding significant harm to their physical and psychological well-being. However, the source of harm differs, and this potentially carries two significant consequences when considering whether climate refugees should have a right to choose their host.

First, the right of political refugees to remain in their host state is temporary. The normatively distinct claim of the political refugee is non-refoulement, which is the obligation of states not to return refugees to the state where they face political persecution (Lister, 2014: 620). This effectively confers on political refugees the right to remain in a safe country only for so long as they are at risk of persecution in their home country. In contrast, climate refugees that require permanent resettlement do not have the option of returning to their home state. This means that choice of host matters more to climate refugees than political refugees, although it clearly matters a great deal for political refugees, especially when the risks of persecution are likely to last a life time.

Second, states that host political refugees typically do not have any connection with, or bear any responsibility for, the political persecution suffered by political refugees in their home state. However, connections between host states and political refugees can sometimes be drawn in the case of military intervention by the host state in the political refugee’s home state. Even well-known communitarians, who uphold the right of communities to determine who should enter and who should stay within their territory, have accepted that a community has a special moral responsibility towards political refugees in those circumstances where the community has contributed to their status as refugees. For example, Michael Walzer has suggested that contributing to their ‘injury’ creates the relevant affinity and ‘thus Vietnamese refugees had, in a moral sense, been effectively Americanized even before they arrived on these shores’ (Walzer, 1983: 49). Whether this obligation on the part of states to receive should be accompanied by a right of refugees to choose is less clear.

Under the Refugee Convention there is no formally acknowledged right for political refugees to choose their host. Asylum seekers who are registered with the United Nations High Commission for Refugees (UNHCR) and languishing in refugee camps awaiting resettlement may request to be resettled in particular states, but states are under no obligation to receive them. The UNHCR refugee resettlement program is merely a voluntary program designed to facilitate burden sharing among signatories to the 1951 Convention, and the decision to accept a particular quota of refugees is purely at the discretion of the host state. Irregular entry is the only means by which political refugees can ‘choose’ their host state, and they do this on the strength of the international customary law principle of non-refoulement, which is binding on all states and also enshrined in the Refugee Convention. However, the practice of offshore processing and detention has enabled host states to take away this de facto ‘choice’, and asylum seekers who are found to be political refugees have no right to choose where they are resettled. The responsibility of states to deal with the plight of political refugees is generally understood as a shared or collective responsibility. In practice, most of the load is borne by developing states since they tend to be closer to the countries generating flows of asylum seekers. Given the significant differences in wealth and capacity between developed and developing countries, the current distribution of refugee load is widely perceived to be an unfair distribution by refugee advocates, and that a fairer allocation of responsibilities should focus on the capacity or capabilities of states to absorb and settle refugees.

In contrast, all states are causally implicated in the plight of climate refugees by authorising or allowing activities within their territory that generate greenhouse gas emissions, albeit in varying degrees. Moreover, the increasing risk of displacement as a result of rising seas and other unavoidable impacts is a direct result of the collective failure of states to address climate change in an effective matter. Prima facie, this ‘contribution to injury’ and the ongoing collective failure to prevent the continuation of the injury with full knowledge of the increasingly severe impacts that will be suffered by vulnerable populations, combined with the permanent character of displacement that is collectively caused, provide strong grounds for a general state duty to receive and a corresponding right on the part of climate refugees to choose their host. This would rest on the recognition that all states are collectively and causally responsible for their plight, albeit in varying degrees but which cannot and need not be disaggregated and measured, and that it would be profoundly unjust to allow concern over the burden of receiving climate refugees to trump their fundamental needs.

Yet while it is possible to draw a morally relevant distinction between political and climate refugees that puts climate refugees in a stronger position than political refugees to claim a right to choose their host state, this is unlikely to be sufficient to engender the political motivation of states to accept such a right.

As we have seen, states have resisted any kind of fault-based or strict liability model or unjust enrichment model of responsibility under the PPP and BPP in the climate negotiations. We have also seen that states are much more favourably disposed to the APP in the real world of climate negotiations and refugee politics, and that this cannot ground a universal right on the part of climate refugees to choose any state they wish since not all states would have the requisite capacity to receive. Finally, we can expect states to be reluctant to accept a responsibility to receive if they consider the application of a common responsibility to receive would produce unfair or politically unpalatable consequences.

One way around this problem might be to restrict or qualify the right to choose in ways that minimise unfair consequences for states. For example, James Souter has argued that the causal connection between a particular state’s action and refugee vulnerability must be strong and ongoing, that a state’s responsibility should only arise when there is a foreseeable connection between the state’s actions and displacement, and that the refugee’s choice should be given considerable weight but may nonetheless be overridden by other factors (Souter, 2014: 179). He applies these principles to both political and climate refugees and defends them as a form of reparation for past injustices. In the case of climate refugees, he suggests that the assignment of reparative responsibilities would implicate only the world’s largest polluters (Souter, 2014: 340).

However, Souter’s case glides over the important differences between political and climate refugees. While some political refugees may be able to establish a strong and ongoing causal connection between the actions of a particular state and the persecution they are experiencing in their home state, climate refugees have a case against all states as we have seen. In any event, conferring a right on climate refugees to single out as their host particular states that are seen to be ‘the most culpable’ or ‘the most capable’ (via unjust enrichment) as a form of just dessert is a singularly unproductive way for them to start a new relationship with their host nation. Nor would climate refugees necessarily wish to single out the most culpable or capable. In the case of the wholesale evacuations of small island nations that wish to retain their right of self-determination along with their control over their submerged territory to which they are entitled under international law, then settling in the regional neighbourhood in order to retain access to their territory may be a bigger concern than the right to choose a major historical emitter or wealthy country. Moreover, island inhabitants might prefer to settle in those neighbouring countries with which they share a greater cultural and/or linguistic affinity. This is not to suggest that all island inhabitants would want to, or have to, make these choices, but it does highlight the fact that what might matter most to the inhabitants of small island states in choosing where to resettle may have little to do with the PPP, BPP or indeed the APP and more to do with geographic proximity and cultural connection. From the standpoint of climate refugees, the need to relocate is a need of necessity that ought not to be overridden by state concerns about relative burden sharing. In these circumstances, a right to choose would provide a form of recognition of, and partial compensation for, the injustice and trauma of their loss and damage.

Yet we still have two major political obstacles to confront. The first would be political resistance by states to a right to choose by climate refugees as a form of reparation for loss and damage. The second would be domestic political fears in many states that they may be overburdened by climate refugees, and this fear would have negative consequences for climate refugees in any efforts to negotiate such rights, or exercise them if granted.

These problems suggest the need for a more creative solution that is able to cater for the special needs of climate refugees while reducing the potential for political hostility by states to the idea of a general responsibility to receive climate refugees and a right to choose. This can be done by severing the connections between the differentiated state responsibility to assist on the basis of the APP, on the one hand, and the common responsibility to receive and the right to choose, on the other. The latter would be based on states’ collective contribution to injury and injustice and would therefore avoid the contentious politics of singling out and blaming, or assigning burdens to, individual states by providing incentives. Whereas the right to choose and the responsibility to receive are necessarily correlative, the responsibility to assist is detachable. This move would enable both receiving states and climate refugees to receive financial and other assistance to defray the costs of resettlement. In the case of receiving states, assistance would be provided on the basis of relative capacity to receive and relative proportion of refugees received. There are significant precedents for resource transfer mechanisms to assist developing countries to meet common legal obligations under particular treaties. For example, the Multilateral Fund set up under the Montreal Protocol to assist developing countries to meet their obligations under the treaty served as a precedent to the Green Climate Fund set up under the UNFCCC to assist developing countries meet their national mitigation and adaptation responsibilities. Likewise, a differentiated responsibility to contribute to the Warsaw International Mechanism on Loss and Damage based on the APP would enable all states to fulfil their common responsibility to receive climate refugees. In short, the case for the ‘right to choose’ is likely to become more politically viable, and the resistance of states to accept their common responsibility to receive them is likely to weaken, the more that climate refugees and their host states are assured of adequate support from states according to states’ differentiated responsibility to assist under the APP. It follows that the first political priority should be to focus on building a robust funding mechanism to assist climate refugees to resettle.

However, while this proposal might soften political resistance by some states it is unlikely to eliminate it, especially if climate refugees move as a large group. Much would depend on domestic politics, including the ability of climate refugee advocates to frame the debate in ways that attract public sympathy. And here there is arguably less cause for public hostility or invasion anxieties due to the absence of many of the factors that have given rise to mistrust, ignorance and hostility to irregular entry by asylum seekers. First, the number of climate refugees is expected to be much smaller than political refugees. On one estimate, the number of people that is expected to be forced to move as a result of the permanent inundation of land by sea under a worst-case scenario is around 15 million people over the 100-year period 2030–2130, allowing for local adaptation (Laczko and Aghazarm, 2009: 88). This is a very small number compared to 16.7 million political refugees plus 1.2 million asylum seekers counted by the UNHCR in one single year in 2013 (UNHCR, 2013: 2). Second, it would be a relatively easy matter to determine who is a climate refugee and they would be less dependent on so-called people smugglers, since they can be expected to travel with the required documentation, especially when it is organised as an orderly movement by the relevant governments. The exception would be emergency evacuation due to sudden onset, climate-related disasters, which is likely to attract public sympathy in host states.

It might be objected that the foregoing defence of the right of climate refugees to choose their host would have the unfortunate effect of converting political refugees into second-class refugees. Indeed, as Luara Ferracioli has argued, the definition of refugees in the 1951 Convention is under-inclusive insofar as there are many kinds of serious vulnerability and hardship that lead to forced migration (Ferracioli, 2014). She also notes that there is no motivation on the part of states to negotiate a new treaty to expand this definition and that if states were to negotiate a new treaty it would be weaker than the 1951 Convention, which creates political dangers for reform initiatives. The same might be said of the UNFCCC’s burden sharing norms, but for the new opportunity presented by the Warsaw International Mechanism.

However, we have seen that there are morally relevant distinctions between political and climate refugees. In addition, climate refugees will not have the cultural possibilities afforded to diaspora populations of political refugees in terms of further access to their own customs, literature, music, and so forth. A right of climate refugees to choose their host would therefore provide both political and cultural justice. In any event, if a policy window has opened for climate refugees, which is not on the horizon for political refugees or other displaced persons, should it be passed over out of solidarity with the plight of a broader class of displaced persons? There is nothing to suggest that the proposed reforms under the Warsaw International Mechanism would make political refugees and other displaced persons worse off, and it is possible that reforms might serve as a positive precedent for this broader class of displaced people in the longer term.

The foregoing proposals would require widespread and skilful advocacy not only by refugee advocates but also like-minded states, and we can expect that many states would favour only a discretionary responsibility to assist and receive, and a right on the part of climate refugees to have considerable weight given to their choice of host, but without any guarantees. SIDS would therefore be wise to take pre-emptive action by seeking to negotiate individual bilateral treaties with their preferred states prior to their forced migration. Yet even these negotiations are also more likely to be successful if it were known that they and their hosts would receive generous assistance under a funding mechanism such as the Warsaw International Mechanism for Loss and Damage.

Conclusion

It should be clear from the foregoing discussion that no single principle of allocating responsibility can provide justice to climate refugees while also treating states fairly. As Caney (2010, 2012) has shown, whatever principles we select, there is always ‘a remainder’, which leads inevitably to some kind of hybrid approach. The PPP leaves unaddressed climate-related harms that cannot be attributed to particular polluters, and it is likely to be a nonstarter in any international negotiations to address loss and damage, including climate-induced migration. Yet while the negotiating history of the climate regime has shown that the PPP has been politically unattractive, it can be expected to play an increasing role in future civil litigation for climate-related damages at the domestic level which may ultimately force many of the structural changes, particularly in the political economy of energy production and use, which most states have so far resisted. It would therefore be politically foolhardy to abandon the liability model.

Likewise, the BPP is limited by the degree to which the beneficiary actually benefited from unwittingly causing a harm, which may not be sufficient to cover the loss suffered by, or assistance required, by the victim. Like the PPP, this principle also faces major measurement challenges. Yet it provides a useful rhetorical bridge between the PPP; unlike, the APP the BPP does not seek to forget history.

The APP is the simplest and most politically attractive principle since it avoids the blame game and is forward looking. Yet in severing the connection between capability and culpability it leads to a discretionary, voluntarist approach by states, who are free to choose their own contribution to loss and damage, just as they do under the existing Refugee Convention, and just as they are doing in preparing their INDCs in the lead-up to signing a new climate agreement in 2015.

The constraints of political feasibility are clearly profound and expose major tensions between what climate justice requires for climate refugees and what the biggest historical emitters and most wealthy states are likely to provide. We have also seen that the dominant moral and political grammar of responsibility as liability remains indispensable but that it is also woefully unsuited to a global risk society. Any attempt to broaden the grammar of responsibility will need to find ways of navigating the inevitable slippage between the agents who have caused and benefited from the problem, those who are culpable and reckless and those with the relevant capacity to provide redress for those who suffer harm and/or prevent the problem from continuing. A new grammar of responsibility dealing with complex interdependence must be matched with practical and workable solutions. The problem is that domestic legal systems, and the current system of sovereign states, are ill equipped to manage incalculable, global risks like climate change. In the longer term, this ultimately points to the need for some kind of global insurance system or Climate Superfund, to which a myriad of actors could be obliged to contribute (including states, corporations, financial institutions and individuals) on a regular basis, and from which a myriad of climate vulnerable communities should be entitled to draw. Meanwhile, the very least that the parties to the UNFCCC should offer climate refugees is sufficient compensation under the Warsaw International Mechanism to enable them to start a new life. And the more that they are able to provide such compensation, the more states are likely to accept the right of climate refugees to choose their host, along with their general responsibility to receive them.