Climate change and displacement: Towards a pluralist approach

Abstract

This paper sets out a research agenda for a political theory of climate displacement, by critically examining one prominent proposal—the idea of a normative status for ‘climate refugees’—and by proposing an alternative. Drawing on empirical work on climate displacement, I show that the concept of the climate refugee obscures the complexity and heterogeneity of climate displacement. I argue that, because of this complexity and heterogeneity, approaches to climate displacement that put the concept of the climate refugee at their centre will fail to treat like cases alike and relevantly different cases differently. In response to these failings, I outline an alternative—the pluralist theory of climate displacement—which confronts the specific challenges that climate displacement poses in different practical and institutional contexts, whilst also treating climate displacement as a unified phenomenon at the second-order level of burden-sharing.

Climate displacement—the displacement of people driven at least in part by the impacts of climate change—has begun to be taken seriously in the international order. In 2016, the United Nations General Assembly (UNGA) adopted the New York Declaration for Refugees and Migrants, which explicitly recognised that people move “in response to the adverse effects of climate change, natural disasters (some of which may be linked to climate change), or other environmental factors” (UNGA, 2016: article I.1). Within the domain of climate politics, the 2010 Cancún Agreements encouraged states to pursue “coordination and cooperation with regard to climate change induced displacement” (UNFCCC, 2011: sec. 13(f)).

At present, these forms of international governance remain mostly aspirational, but they reflect the fact that climate displacement is now firmly on the international agenda. As the impacts of climate change unfold, the political salience of climate displacement will only grow. And although there is significant disagreement over exactly how much displacement will take place as a result of the impacts of climate change, there is now a consensus, reflected in the judgment of the Intergovernmental Panel on Climate Change that climate change impacts are likely to lead to an increase in displacement over the course of the 21^st century (IPCC, 2014: 73). This creates both an opportunity and an imperative for political theorists to examine the duties that we owe to those displaced by the impacts of climate change.

Despite growing literatures on both climate justice and migration justice, political theorists have paid relatively little attention to climate displacement. Most of the literature on the topic has focused on the case of small-island states threatened by sea level rise (Stilz, 2019; Eckersley, 2015; Ödalen, 2014; Kolers, 2012; Nine, 2010; Risse, 2009). But as we will see, this case is only one part of the normative landscape of climate displacement. In the small literature in political theory that does aim to provide a more comprehensive analysis of climate displacement, one idea has enjoyed significant popularity: the idea of a special normative status for ‘climate refugees.’

In this paper, I argue that proposals for a normative status for climate refugees—which I call monist approaches—are misguided. By focusing on the concept of the ‘climate refugee,’ they treat those displaced by climate impacts as an internally homogenous group who can be readily distinguished from those displaced by causes other than climate change. The empirical literature on climate displacement, however, tells us that climate displacement is heterogeneous—it takes many different forms in different contexts—and causally complex—climate impacts are usually only one part of a constellation of interrelated causal drivers. This matters, so I argue, because it means that proposals that focus on an idealised climate refugee will predictably exhibit two forms of morally troubling arbitrariness: they will fail to treat like cases alike and relevantly different cases differently.

In response to the failings of the monist approach, I outline an alternative, which I call the pluralist theory of climate displacement. The pluralist theory has two main commitments. First, it disaggregates climate displacement, focusing on the problems that it poses in particular institutional and practical contexts. Second, it integrates climate displacement with other, non-climatic forms of displacement, by making first-order responses independent of the climatic credentials of an individual's displacement. By disaggregating and integrating climate displacement, rather than focusing on an idealised climate refugee, the pluralist theory escapes the problems that afflict the monist approach.

There is, however, an important moral difference between climate displacement and other, non-climatic forms of displacement. Climate change raises distinctive questions about responsibility which do not arise—or which arise only in importantly different ways—in relation to other forms of displacement. The pluralist approach accommodates these considerations of responsibility not at the level of first-order responses to displacement, but at the second-order level of burden-sharing.

The theory that I set out here remains only a sketch. This is because, according to the pluralist theory, different principles are appropriate for different contexts of climate displacement. As such, a full analysis of climate displacement would require a detailed investigation of the normative terrain of each context. The sketch that I provide does, however, set out a research agenda for a political theory of climate displacement.

In broader terms, the aim of this paper is to dislodge a picture that holds great sway in public and academic discussions of climate displacement: the figure of the ‘climate refugee.’ The figure of the climate refugee by now occupies a significant place in our moral imagination. In François Gemenne's words, climate refugees have become “the human face of climate change” (Gemenne, 2011). My suggestion is that the figure of the climate refugee is a ‘picture’ in the Wittgensteinian sense: it has “held us captive” (Wittgenstein, 2007: 53) and prevented us from adopting alternative perspectives on climate displacement. ¹ This picture of climate displacement restricts our vision—it puts us in a situation of “aspectival captivity” (Owen, 2002: 217–19)—and so impedes our moral reflection. By dislodging this picture, I hope to enable us to start afresh and look at climate displacement in a more contextually specific way.

The monist approach

In the nascent literature on climate displacement, a number of proposals have been set out which propose to create a special normative status for ‘climate refugees.’ This status would grant those identified as climate refugees a particular set of rights and would imply a set of correlate duties on the part of the international community. These accounts usually take the form of proposals for a new international institution, usually in the form of a multilateral treaty for climate refugees. I call these approaches monist approaches.

The most influential example of the monist approach comes from Biermann and Boas (2010; 2008), who have outlined a proposal for a new legal instrument that would be constituted as a Protocol to the United Nations Framework Convention on Climate Change (UNFCCC), called the Protocol for the Recognition, Protection and Resettlement of Climate Refugees. Biermann and Boas (2010: 75–76) argue that the Protocol should identify a group of “climate refugees” whose treatment should be regulated by five core principles of governance: (i) the principle of planned relocation and resettlement; (ii) the principle of resettlement instead of temporary asylum; (iii) the principle of collective rights for local populations; (iv) the principle of international assistance for domestic measures; and (v) the principle of international burden-sharing. Taken together, they see these principles as articulating an account of what the international community owes to those displaced by climate impacts. The core rights to which climate refugees would be entitled, on Biermann and Boas's view, are rights to collective resettlement, either within their state or internationally. As part of the proposal, Biermann and Boas (2010: 75, 79–82) also outline a funding mechanism, which operates according to a grant system where developed countries contribute proportionally to funding to a shared pool according to their moral responsibility for the plight of climate refugees.

Other structurally similar proposals have also been put forward. For example, Docherty and Gianni (2009) have called for a sui generis legal convention for “climate change refugees.” This convention, they argue, should guarantee climate refugees access to a range of rights, including both rights specifically related to movement and various civil, political, social, cultural and economic rights (Docherty and Giannini, 2009: 376–79). Another proposal has been put forward by Byravan and Rajan (2010; 2015), who argue for a special right to free movement for “climate exiles.” On their view, ‘climate exile’ status should be inscribed in an international treaty or protocol to the 1951 Refugee Convention, and should entitle its bearers to an individual right to migrate to and settle in a safe country. Historically high-emitting states would have primary responsibility for “providing immigration rights” for climate exiles (Byravan and Rajan 2010: 253).

These proposals all differ significantly in their details. They identify different groups of persons as climate refugees (or ‘climate exiles’), specify different rights and correlate duties associated with climate refugee status, and take different institutional forms—which may be more or less revisionary when it comes to existing legal practice. But despite these differences, they share an important feature: they take the task of addressing climate displacement to require that we pick out a group of ‘climate refugees’ as a category of concern and articulate a set of rights to which persons in that group are entitled. In what follows, I argue that it is this core feature of the monist approach that means that it gets off on the wrong foot.

Before making this argument, however, it is worth noting why the idea of a normative status for climate refugees appears attractive, at least at first glance. The motivation behind the proposal is that it appears well-equipped to address what is sometimes called, in the parlance of international legal and humanitarian practice, a “protection gap” (Türk and Dowd, 2014) in relation to climate displacement. A protection gap refers to the idea that existing legal frameworks fail to offer protection to some people who have justified claims to protection. In the case at hand, the existing legal frameworks that govern displacement in the international order fail to offer protection to those displaced by climate change impacts: they fall through the ‘gaps’ in international protection. For example, those displaced by climate change impacts are not eligible for protection under the 1951 Refugee Convention, which restricts protection to those who suffer a ‘reasonable fear of persecution’ on the basis of certain protected characteristics. ² Similarly, many of those displaced by climate change impacts will not be eligible for protection under the regimes of governance for the protection of ‘internally displaced persons’ (IDPs) (Koser, 2011). And the fragmented and largely ad-hoc system of ‘complementary protection’ for those who fall outside the scope of these regimes provides little comfort for those seeking firm guarantees of protection (McAdam, 2012: 52–99).

Existing legal frameworks fail to provide robust forms of protection for those displaced by climate change impacts. Clearly, the judgement that this amounts to a ‘protection gap’ depends upon an assumption that those displaced by the impacts of climate change ought to be protected under legal frameworks in some way, but I do not think that this judgement is particularly controversial. Proponents of the monist approach see their proposals as a way of closing this protection gap (Biermann and Boas, 2010: 72–74; Docherty and Giannini, 2009: 357–61). In my view, they are right to point out that the current international legal order fails to protect those who have compelling claims to protection. It is only their proposed solution, the creation of a normative status for climate refugees, that I take to be mistaken. In order to see why, it is useful to look a little more closely at the concept of the climate refugee.

The concept of the climate refugee

Talk of climate refugees is ubiquitous in popular, policy and academic discourses. Within the monist approach, however, the concept of the climate refugee plays a specific role: it ascribes a particular normative status, with a corresponding set of rights and correlate duties, to those that it picks out.

Climate refugees have been defined in different terms in different versions of the monist approach. Some use the term to refer to people displaced by climate impacts in general: Docherty and Gianni (2009: 361), for example, target those who flee “as the result of sudden or gradual environmental disruption that is consistent with climate change and to which humans more likely than not contributed.” Others use it to refer to groups of people displaced by specific climate impacts: Biermann and Boas (2010: 64), for example, restrict their definition of climate refugees to those fleeing from three “largely undisputed” climate impacts: sea level rise, extreme weather events, and drought and water scarcity; and Byravan and Rajans (2010: 252) proposal targets those facing permanent displacement due to the loss of habitable land. Though each of these definitions identifies a different group of people, each proposal puts the concept of the climate refugee at the centre of its project.

The concept of the climate refugee, however, faced significant pushback in migration studies (Suhrke, 1994; Black, 2001). Critics have argued that the concept of the climate refugee depends on untenable monocausal assumptions about the relationship between climate change and displacement and fails to distinguish between proximate and underlying causes of displacement—problems that “strike to the core of the literature on environmental refugees” (Black, 2001: 3). This does not mean that climate change is unimportant in driving displacement. Indeed, there is now a wealth of research that documents the ways in which climate change impacts interact with displacement (Piguet, Pécoud and de Guchteneire, 2011). But the empirical literature tells us that climate displacement is both complex and heterogeneous.

The complexity of climate displacement refers to the idea that climate change impacts interact with other factors in ways that make it difficult to isolate climate change as the cause of any particular instance of displacement. Empirical research on climate displacement shows that climate impacts are usually intertwined with other social, economic, environmental, political, and demographic drivers of displacement, and are mediated by social facts about vulnerability and resilience (Black, Kniveton et al., 2011; Black, Adger et al., 2011; McLeman and Hunter, 2010; McLeman and Smit, 2006). The heterogeneity of climate displacement refers to the fact that climate displacement takes different forms in different contexts. The empirical literature identifies a variety of different ways in which the impacts of climate change can affect patterns of displacement (Kälin, 2010; Piguet et al., 2011). Extreme weather-events destroy homes, environmental degradation undercuts the viability of livelihoods, sea level rise and coastal erosion force communities to relocate, and risks to food and resource security magnify the sources of political instability, and more.

To illustrate these features of climate displacement, consider a few examples. First, take displacement in the aftermath of Hurricane Katrina, an extreme weather event of the type that is likely to become more frequent and intense as climate change accelerates. As Katrina unfolded, its effects were mediated by pre-existing vulnerabilities structured along the lines of race and class. Evacuation messages were less likely to reach, or to be trusted by, low-income and Black residents in New Orleans, many of whom remained behind because they were concerned that public authorities would not protect their property, or because they provided care for someone who was unable to leave (Bolin and Kurtz, 2007). The evacuation plan for the city depended on residents being able leave by car, but rates of car ownership were much lower amongst Black and low-income communities (Bullard, 2008). As a result, the impacts of the disaster fell most heavily on low-income and Black residents.

Second, consider out-migration from Mexico to the USA. At the aggregate level, statistical analyses show that reductions in crop yields relating to climate change have an important causal role in driving migration from Mexico to the USA (Feng et al., 2010). Importantly, though, other factors also have also been identified as having an influence on who moves from Mexico to the USA at the aggregate level, including the relative growth in labour supply over labour demand in Mexico (Hanson and McIntosh, 2009), structural transformations in the Mexican economy (Massey, 1987), and changes in US immigration policy (Reyes, 2004). So far as we know, any given case of out-migration from Mexico may result from any one or a combination of these factors—as well as, of course, decisions taken by migrants themselves.

Third, consider the relocation of Native communities in Alaska. In coastal Alaska, the village of Shishmaref has been at risk of coastal erosion since the 1950s, but this risk has been exacerbated in by thinning ice, shoreline erosion, severe storms and permafrost exposure relating to climate change (Marino, 2012). The inhabitants of Shishmaref are members of the Iñupiaq tribe, whose daily lives are intimately bound up with the land they occupy. The residents of Shishmaref are seeking relocation within tribal territory, but the relocation process has stalled. Distrust between tribal and federal authorities has marred the processes, and funding has been an important obstacle: until recently, legislation governing the activities of the Federal Emergency Management Agency (FEMA) provided only for post-disaster rebuilding of homes in their original location after a presidential declaration of a disaster (Bronen and Chapin, 2013).

In each of these cases, displacement is complex in the sense that multiple factors interact together to produce the displacement (or immobility) outcome. Often, factors that affect displacement outcomes fall along classic fault lines such as class, race, indigeneity and gender. But other factors also influence displacement outcomes, including the legal and policy context, the structure of labour markets, and aspects of the built environment. The difference between the cases also illustrates the heterogeneity of climate displacement. These cases differ in a number of ways, and there are a number of axes according to which we can distinguish them: slow-onset vs. rapid-onset, internal vs. international, reactive vs. proactive, community-level vs. individual-level, and so on.

Critics have argued that we should reject the concept of the climate refugee because it obscures these features of climate displacement. Mike Hulme, for example, has objected to Biermann and Boas's proposal—which excludes cases where “secondary, or indirect, effects of climate change” lead to displacement and expects scientific assessors to be able to identify climate refugees (Biermann and Boas, 2010: 64, 77–78)—on the basis that their usage of the concept of the climate refugee “implies a monocausality about the reasons for migration that just does not exist in reality” (Hulme, 2008: 50). Etienne Piguet, Antoine Pécoud and Paul de Guchteniere have argued that the language of climate refugees is unhelpful and that it is necessary “to disentangle the different kinds of mobility that may be connected to environmental factors” (Piguet et al., 2011: 14).

At least at first glance, however, it is not clear why this should lead us to reject the monist approach. It is true that the concept of the climate refugee obscures these features of climate displacement. It singles out one cause of displacement—the impacts of climate change—out of many that are entangled, and it groups together all of those whose displacement can be linked to that cause, regardless of any internal differences within that group. But the purpose of the concept of the climate refugee, at least as it figures within the monist approach, is not to provide an accurate description of the dynamics of climate displacement. Rather, it is to establish a certain set of rights and duties to those granted recognition as climate refugees. Just as the concept of corporate personhood is better understood in terms of its role in establishing certain rights and duties (such as the liability of corporations to be sued in court) rather than in terms of its representing truths about corporate agency, so too is the concept of the climate refugee better understood in terms of its role in establishing rights and duties rather than in terms of its faithfully representing the dynamics of climate displacement.

My suggestion is that the monist approach can nonetheless be rejected on the grounds that the concept of the climate refugee obscures the complexity and heterogeneity of climate displacement. This is not because idealising away these features of climate displacement is inherently objectionable, but because rather because an approach that ignores the complexity and heterogeneity of climate displacement will predictably fail to treat like cases alike and relevantly different cases differently.

The arbitrariness of the monist approach

My principal objection to the monist approach is that its focus on the concept of the climate refugee means that it would predictably exhibit two morally objectionable forms of arbitrariness: it would fail to treat like cases alike and relevantly different cases differently. The idea that our institutions ought not to distinguish between cases on grounds that are arbitrary from a moral point of view is a basic principle of formal equality, which I take to be relatively uncontroversial. One characteristic expression of this principle is Rawls's (1999: 5) claim that “institutions are just when no arbitrary distinctions are made between persons in the assigning of basic rights and duties.” Where there are no morally relevant differences between cases, they should be treated alike; where there are morally relevant differences between cases, they should be treated differently—in ways that respond appropriately to those morally relevant differences.

My claim is that the monist approach will predictably create morally objectionable forms of arbitrariness in practice. But of course, all legal and institutional frameworks are likely to be over- and under-inclusive in practice, and this problem can hardly be avoided entirely (Dempsey and Lister, 2017: 217). So, my suggestion is not that any element of arbitrariness in practice means that we should reject a proposal. Rather, my suggestion is that the monist approach has a particular propensity to create morally objectionable forms of arbitrariness, because of the way it conceptualises climate displacement. The problem is not that the monist approach is unable to accommodate borderline or hard cases. It is rather that it is unable to accommodate typical cases of climate displacement. This problem is not an inevitable feature of any approach to climate displacement and can be avoided (or at least significantly attenuated) by adopting an alternative approach that does not depend on the concept of the climate refugee. We can take each form of arbitrariness in turn.

First is the claim that the monist approach would fail to treat like cases alike. There is a strong and a weak version of this claim. The strong version of the claim says that whether or not someone's displacement is caused by climate change should have no bearing on the rights to which they are entitled—that it is arbitrary form a moral point of view. Jane McAdam, for example, has argued that “a ‘climate refugee’ treaty would privilege those displaced by climate change over other forced migrants (such as those escaping poverty), perhaps without an adequate (legal and/or moral) rationale” (McAdam, 2012: 187–88; see also Cole, 2018). Under the monist approach, those displaced by climate impacts would be guaranteed international protection, whilst those whose displacement has causes unrelated to climate change, but who are otherwise in a similar position, would not. Unless there is something special about those whose displacement is caused by climate change, this appears to be morally arbitrary.

There is a sense in which this strong claim is correct: those displaced by climate impacts do not appear to have a stronger claim to first-order forms of protection than those displaced by non-climatic drivers of displacement such as conflict, economic insecurity or non-climatic natural disasters. But even so, this strong claim does not undermine the monist approach. This is because there is nothing in the monist approach itself that precludes others with similar needs from being granted similar forms of protection and assistance under other, complementary institutional frameworks, at least in principle.

There is also at least one morally relevant reason to distinguish between climate displacement and other forms of displacement—namely that climate displacement raises distinctive questions of responsibility which may not arise (or arise only in different forms) in other contexts (Buxton, 2019; Draper, 2019). This may not bear directly on the first-order question of what we owe to particular displaced persons, but (as I will suggest below) it does have an important bearing on the second-order question of how the costs of protection should be distributed. This might provide a rationale for developing separate institutional frameworks for climatic and non-climatic forms of displacement.

The weaker version of the claim that the monist approach will fail to treat like cases alike is a more minimal, internal criticism, which accepts (at least for the sake of argument) that it is justifiable to distinguish between those displaced by climate impacts and those displaced in other ways. But it points out that the complexity of climate displacement means that, in practice, decisions about who counts as a climate refugee will inevitably involve significant arbitrariness.

The monist approach requires us to be able to identify those who count as a climate refugee and those who do not. As we have seen, different proposals set out different definitions of the climate refugee, with some focusing on specific climatic parameters and some focusing on climate change in general. But whichever they focus on, they require us to be able to distinguish between different people on the basis of the cause of their displacement. The problem with this is that the empirical finding that climate displacement is complex—that climate impacts interact and are intertwined with other drivers of displacement—means that it will often be difficult to differentiate those whose displacement is caused by climate change from others.

This problem has both an epistemic and an ontological aspect. Under the epistemic aspect, the problem is that we may not be able to know when climate change is appropriately singled out as the cause of displacement. The causal chains that lead from climate change impacts to displacement are messy, and it may well be beyond our capacities to untangle these threads of causation. Under the ontological aspect, the problem is that it simply may not be meaningful to identify climate change as a cause of displacement which is isolable from other causes, at least in many cases. The links between climate change and other drivers of migration are often non-linear, and it is not clear that combined causes in these cases can simply be broken down into their constituent parts (Black, Kniveton et al., 2011).

As McAdam (2012: 197–98) has pointed out, the complexity of climate displacement means that any attempt to identify climate refugees would face serious hurdles. Decisions about who to count as a climate refugee will often not depend on whether one's displacement really is caused by climate change, because there is often no good answer to that question. At best, decisions are likely to be made on the basis of how salient the climatic drivers of displacement appear to decision-makers in comparison to non-climatic drivers. In many cases, especially where climate impacts are slow-onset, the climatic aspects of displacement will be largely invisible and other, more proximate causes—such as civil conflict or labour market pressures—will be more obvious (McAdam, 2012: 36–38). At worst, decisions may be made on the basis of the political expediency of making claims about the causal role of climate change. It may suit political actors to lay the blame for displacement on climate change, rather than other drivers (such as conflict) or vice versa. Indeed, we have already begun to see causal claims about climate displacement being mobilised in this way (Betts and Pilath 2017).

This problem appears particularly pressing when it leads to under-inclusion, since this involves denying protection those with justified claims. Unfortunately, under-inclusion appears particularly likely, because the complexity of climate displacement makes it easy for powerful states to shirk responsibility by downplaying the causal role of climate change. But over-inclusion may be a problem too, either for reasons of fairness between claimants seeking scarce protection resources, or because a tendency towards over-inclusion may make states reluctant to sign on to the proposal in the first place. In any case, these reasons of salience and political expediency are clearly not morally relevant reasons for distinguishing between cases. The upshot of this is that, given the complexity of climate displacement, there will be a significant problem of arbitrariness in decisions about who counts as a climate refugee.

The second way in which the problem of arbitrariness manifests in the monist approach is in its failure to treat relevantly different cases differently. This problem arises because its focus on the concept of the climate refugee obscures the heterogeneity of climate displacement. An approach centred on the concept of the climate refugee treats those that it picks out as an internally undifferentiated group, whose members are all entitled to the same set of rights.

Consider the group of people identified as climate refugees under Biermann and Boas's proposal: “people who have to leave their habitats, immediately or in the near future, because of sudden or gradual alterations in their natural environment related to at least one of three impacts of climate change: sea-level rise, extreme weather events, and drought and water scarcity” (Biermann and Boas, 2010: 67). They argue that those in this group should have access to planned, voluntary, and permanent resettlement (Biermann and Boas, 2010: 75). It is not clear at all, however, why all of those who fall within this category should all be entitled to the same set of rights to planned, voluntary and permanent resettlement. There are important differences between those who face displacement due to sea level rise, extreme weather events and drought and water scarcity. Voluntary relocation and resettlement may be appropriate for at least some of those facing threats to their homes due to sea level rise. But those who face the impacts of recurring droughts may well be better served by a programme of circular or seasonal labour migration that allows them to diversify their household's sources of income, for example, rather than by a programme of resettlement. Similarly, those facing displacement due to extreme weather events may be better served by a programme of disaster risk reduction and/or forms of post-disaster relief and recovery assistance, rather than by relocation projects. It is arbitrary to grant each of these groups of persons the same set of rights when there are morally relevant differences between them. This point does not only apply to Biermann and Boas’ proposal, it generalises. Any approach version of the monist will be inattentive to morally relevant differences between different cases of climate displacement, insofar it focuses on the concept of the climate refugee, which obscures the heterogeneity of climate displacement.

A similar criticism of proposals for a climate refugee treaty has been made by McAdam, who draws on fieldwork in the different contexts of Bangladesh, Tuvalu and Kiribati to argue that “a universal treaty may be inappropriate in addressing the concerns of particular communities” and that what is needed is a system of governance that “take[s] into account the particular features of the affected population, in determining who should move, when, in what fashion, and with what outcome” (McAdam, 2012: 188). What is often not appreciated, however, is that this is a moral objection to the monist approach. It is a moral objection because it is a complaint about particular rights and duties being misallocated. For example, consider someone living in a low-lying coastal area threatened by sea level rise who is seeking relocation within her state. On Byravan and Rajan's proposal—according to which those facing threats from sea level rise, including in deltaic regions, count as “climate exiles”—she would be granted “immigration rights” abroad (Byravan and Rajan, 2010: 251–53). Not only is this right not warranted by or appropriate for her circumstances, but it is also a right that is denied to others who may see it as valuable—for example, those who wish to immigrant in order to pursue economic opportunities. At the same time, she is not granted a right to which she appears to be entitled—the right to material and technical assistance in relocating to a safe environment within her state—which may be offered to others who are in a relevantly similar position, such as those living in the shadow of an active volcano. The monist approach would systematically misallocate the various rights and duties that are appropriate for different cases of displacement, because it ignores the heterogeneity of climate displacement.

The monist approach, then, would exhibit two forms of moral arbitrariness. First, it would fail to treat like cases alike, because it ignores the complexity of climate displacement. Second, it would fail to treat relevantly different cases differently, because it ignores the heterogeneity of climate displacement. This problem arises because the monist approach puts the concept of the climate refugee at its centre, and so treats those displaced by climate impacts as a homogenous group who can be readily identified based on the cause of their displacement.

To be clear, this objection is not an objection to the idea of a treaty for climate displacement per se, even though most proponents of the monist approach suggest that the institutional form their proposals should take is something like a multilateral treaty. A modified proposal for a treaty for climate displacement could, in principle, avoid these objections. In order to do so, however, such a treaty would need to disaggregate climate displacement and integrate it with other, non-climatic forms of displacement. In the next section, I sketch a pluralist theory of climate displacement, which puts these two commitments at its centre.

Towards a pluralist theory of climate displacement

The failings of the monist approach stem from the way in which it centres around an idealised picture of the climate refugee, which obscures the complexity and heterogeneity of climate displacement. In order to avoid these problems, my suggestion is that we should dislodge the picture of the climate refugee and focus instead on the specific practical and institutional contexts in which climate change and displacement interact. In this section, I briefly sketch what I call a pluralist theory of climate displacement, which does just that.

The pluralist theory has two core commitments. First is that climate displacement should instead be integrated with other, non-climate forms of displacement. No-one's fate should hang on climate impacts being identified as the cause of their displacement, given that climate impacts will most often be intertwined with other drivers of displacement in complex ways. Second is that climate displacement should be disaggregated, rather than being viewed as one, undifferentiated phenomenon. Different principles and institutional forms will be appropriate for the different contexts in which climate change and displacement interact. In each context, the task of the theory is to identify the problems that climate displacement creates for our institutions and practices, and to propose ways of reforming those institutions and practices. ³ This is what we might call a problem-driven approach to climate displacement.

Contexts of climate displacement

Here, I identify five contexts in which climate change presents problems for our existing institutions and practices: community relocation, territorial sovereignty, labour migration, internal displacement, and the refugee regime. ⁴ The particular contexts that I identify are unlikely to be exhaustive, given the unpredictable nature of climate change impacts. Nor should they be viewed as mutually exclusive categories for classifying each person whose displacement is related to climate change. Rather, they are best viewed as focal points that serve to orient our reflection. In the space available to me here, I can only provide a preliminary assessment of the problems that arise in these different contexts. This can be viewed as the statement of a research agenda for a pluralist theory climate displacement.

First is community relocation. Community relocation is the practice of relocating the inhabitants of a community in response to threats that make their environment unsuitable as a site for human occupancy. It is broader than climate-induced community relocation: for example, community relocation often takes place in the context of development projects that make certain sites uninhabitable (see Penz et al., 2011), or in response to non-climatic hazards such as earthquakes or volcanos. In the context of climate change, impacts such as sea level rise, recurrent flooding, and landslides can all create risks that make it difficult for established communities to remain where they are. The nature of these impacts means that are important disanalogies with development-induced displacement: for example, development-induced displacement raises questions about when, if ever, public benefits can justify relocation projects, whereas in the case of climate change, there is often little prospect of stopping the impacts that necessitate relocation (Drydyk, 2013). At the same time, it is rare that climate impacts alone necessitate relocation: the impacts of climate change can often be attenuated by adaptation. This means that questions of procedural justice, concerning who has the power to decide when, where and how a community relocates, are often particularly salient (Draper and McKinnon, 2018).

Second is territorial sovereignty. The habitability of small-island states such as Kiribati, Tuvalu, the Maldives, Vanuatu and the Marshall Islands is threatened in the long term by climate impacts such as sea level rise and coral bleaching (Barnett and Adger, 2003). In this context, there is a threat of what Vaha (2015) has called “state extinction”: climate impacts threaten the land base that enables the exercise of territorial sovereignty in small-island states, at least under the current principles that regulate statehood in the international order. This presents a novel challenge to the practices of statehood and sovereignty, which are the principal way in which the value of self-determination is realised (however imperfectly) in the contemporary international order. The case of small-island states requires us to think through the ways in which we might reconfigure the practices of statehood and sovereignty in order to protect small-island peoples’ rights to self-determination. This has already been a significant focus of attention amongst theorists of territorial rights, who have made proposals such the redistribution of territory between states (Nine, 2010; Angell, 2021), models of ‘de-territorialised’ statehood (Ödalen, 2014), and forms of intrastate autonomy (Stilz, 2019: 177–85).

Third is labour migration. Some climate change impacts, such as thawing permafrost, salination of ground water, desertification, and recurring floods or droughts, can gradually degrade the environmental background conditions upon which people rely for their livelihoods or subsistence. These climate impacts typically interact with drivers of migration such as labour market pressures and existing forms of poverty and vulnerability. The evidence suggests that labour migration can function as a mechanism of climate change adaptation at the household level (Black, Adger et al., 2011). This raises questions concerning whether states may permissibly use labour migration policy as a tool of adaptation, what terms they may permissibly impose on would-be labour migrants, and whether those facing climatic threats to their livelihoods or subsistence have claims forms of in situ adaptation to help them remain where they are (Draper, 2021a; Chang, 2009). The context of climate change alters the moral landscape of debates about labour migration: instead of conceiving of labour migration from poor to rich states as an agreement for mutual advantage or a tool for alleviating global poverty, it now appears as mechanism of adaptation to climate change.

Fourth is internal displacement. Sudden-onset extreme weather events associated with climate change, such as flooding, typhoons, hurricanes and wildfires, can trigger displacement. Most people displaced in this way are likely to remain within their state's borders, and so to count as IDPs under international legal frameworks (Koser, 2011). IDP status, like refugee status, is a normative status that implies a set of rights that its bearer can claim and correlate duties that others owe. Compared to refugee status, however, IDP status has received little attention in political theory (for recent exceptions see Draper, 2021b; Beaton, 2020). Climate change also presents novel challenges to the practices of IDP protection. The regime of protection for IDPs generally presumes that the causes of displacement are internal to the state in question and treats any international assistance in IDP protection as a matter of state discretion (Phuong, 2005). In the context of climate change, however, the idea that internal displacement is a purely domestic affair can no longer be maintained, and so we need an account of the duties of the international community to assist in IDP protection.

Fifth, and finally, is the refugee regime. Under international law, refugees are those who have a ‘well-founded fear of persecution’ on the basis of certain protected characteristics. In some cases, climate change impacts may magnify existing sources of political instability and indirectly contribute to situations in which such persecution takes place, for example where resource scarcity or environmental degradation play a role in fomenting social conflict—although there are almost always existing sources of conflict at play in such cases (Barnett and Adger, 2007). But beyond this, there are likely to be some people who are displaced in refugee-like situations at least in part due to climate impacts who do not qualify as refugees. Some political theorists have argued that the basis for refugee status should be expanded (Shacknove, 1985; Gibney, 2004). If they are right, then those fleeing their state for other reasons, such as the failure of their state to adequately respond to extreme weather events, may also have a claim to refugee status. But even if they are wrong, we still need to know to articulate what we owe to those in refugee-like situations, perhaps under the auspice of some more robust kind of complementary protection status (Lister, 2019).

Each of these contexts of climate displacement raises distinctive normative questions. By disaggregating climate displacement, rather than treating those displaced in the context of climate change as a homogenous group, we can examine the normative questions that arise in each of these contexts on their own terms. In doing so, we need not identify any particular individual as a ‘climate refugee.’ Rather, we should seek to reform the institutions and practices that govern displacement more broadly, such that they are able to accommodate the novel challenges raised by climate change.

Stated at this level of generality, the pluralist theory is compatible with multiple substantive views on what is owed to the displaced in each of these different contexts, and on how our institutions should be arranged to enable those duties to be discharged. But it is worth noting that on the pluralist theory, there is no assumption that the international community itself is the primary bearer of first-order duties in contexts of climate displacement. In some contexts, such as in relation to refugee movement and territorial sovereignty, the international community itself may have first-order duties to the displaced. But in other contexts, such as community relocation and internal displacement, the primary duty-bearer is likely to be state in which the displaced person finds herself. On the pluralist theory, the duties that members of the international community owe in virtue of their contributions to climate change, which I examine in the next section, are second-order duties to assist those who are best placed to discharge first-order duties to the displaced.

The pluralist theory avoids the problems of the monist approach whilst at the same time retaining its principal advantage. The main attraction of the monist approach was that it could close the ‘protection gap’ in the international order relating to climate displacement. The pluralist theory also enables us to close that protection gap, but it does so in a different way. Where the monist approach would close the protection gap through a blanket approach that creates one status for all those displaced by climate impacts, the pluralist theory directs us towards a more bottom-up approach. By reforming our institutions and practices in a range of contexts in which climate change and displacement interact, we can make those institutions and practices suitable for addressing climate displacement, and thus close the protection gap in relation to climate displacement.

The pluralist theory also avoids the two kinds of moral arbitrariness that are exhibited in the monist approach. It is able to treat like cases alike, because no one's fate hangs on their being identified as a climate refugee. This means that no one will be excluded on the basis of morally arbitrary reasons, such as that the climatic drivers of their displacement are not salient to decision-makers. At the same time, the pluralist theory is able to treat relevantly different cases differently, because it disaggregates climate displacement. This means that different principles can apply to different contexts of climate displacement, where there are morally relevant differences that our theory should reflect.

On the pluralist theory, first-order responses to displacement do not depend on whether or not an individual's displacement can be attributed to climate change. But this does not mean that climate change is unimportant in theorising about displacement. Climate change does make an important difference to our theorising about displacement, but its relevance is at the second-order level of burden-sharing, to which I now turn.

Responsibility and burden-sharing

One way that climate displacement differs from other forms of displacement is that it raises distinctive questions of responsibility. Ordinarily, we think of our duties to the displaced as duties of rescue (Gibney, 2004) or as duties arising from our shared participation in a system of sovereign states (Owen, 2020). In the case of climate change, however, we are implicated in the plight of the displaced through our contributions to climate change. Of course, there are also other cases where particular agents can be held responsible for displacement—where belligerents in wars or development actors cause displacement, for example. But in those cases, we can usually draw a fairly clear line from cause to effect, and hold the relevant agents liable to bear the costs of addressing displacement. In the case of climate change, by contrast, the link between our actions and displacement is much more diffuse, such that it is difficult to untangle the causal threads.

This appears to create a problem for the pluralist theory. Since the pluralist theory is committed to integrating climate displacement with other, non-climatic forms of displacement, it appears unable to accommodate these considerations of responsibility. Here, however, I argue that the pluralist approach can accommodate responsibility. Accommodating responsibility requires that we shift our focus from the first-order question of what we owe to those displaced by climate impacts to the second-order question of how the costs of climate displacement should be shared. We can hold contributors to climate change liable to bear the costs associated with the aggregate burden of displacement risk that they impose at the macro-level, rather than holding them liable to bear the costs associated with the displacement of particular individuals at the micro-level.

Proponents of the monist approach have argued that one virtue of their proposals is that they enable us to hold contributors to climate change responsible for climate displacement (Biermann and Boas, 2010: 76; Byravan and Rajan, 2010: 242–53; Docherty and Giannini, 2009: 379). But the complexity of climate displacement undercuts this claim. To see this, note that ordinarily, when we make judgements about responsibility, we identify discrete harmful outcomes caused by discrete actions or series of actions, and assign responsibility on the basis of an identifiable connection between the two. This is the basic model of responsibility attribution that underlies the practices of tort law, for example (Perry, 1992). This tort model generally works well enough as the basis for making judgements of responsibility in our day-to-day lives. This is also the model that the monist approach assumes: it assumes that we can identify particular climate refugees and hold high-emitters liable to bear the costs associated with their protection.

The problem, however, is that the tort model is unable to account for the complexity of climate displacement. The complexity of climate displacement means that we often cannot identify particular individuals whose displacement is caused by climate change with any degree of confidence. Part of the problem here is that the complexity of climate displacement makes it difficult to identify particular individuals displaced by climate impacts. If we cannot reliably identify particular individuals as climate refugees, then we cannot reliably hold high-emitters liable to bear the costs of their displacement, for the very same reasons. Another part of the problem is that in some cases, other parties apart from high-emitters may also bear some responsibility on the tort model. For example, where states fail to take reasonable precautions to avoid displacement, this might reduce the share of the costs for which high-emitters are liable, as under the doctrines of “contributory” or “comparative” negligence in tort law. The complexity of climate displacement, however, makes it very difficult to make these sorts of judgements about responsibility with any degree of confidence.

Instead of adopting a tort-like model of responsibility, in the case of climate displacement we should adopt an insurance model of responsibility (see also Thornton, 2018; Penz, 2010). On this approach, we do not seek to establish responsibility for the displacement of particular individuals by the impacts of climate change. Instead, we see the relevant ‘outcome’ for which high-emitters are responsible as the increased level of displacement risk imposed by climate change. In this way, we can hold contributors to climate change liable for the overall costs that they impose on the regimes of governance for displacement.

The basic idea of this approach is helpfully illustrated with an analogy to workers’ compensation insurance. In late nineteenth century Germany, demands grew for a mechanism to compensate workers involved in industrial accidents without the need to establish fault on the part of employers. Doctrines such as the ‘fellow-servant’ rule (an employer is not liable if a co-worker contributed to the injury) and the ‘contributory negligence’ rule (an employer's liability is diminished to the extent that the employee contributed to their own injury) heavily favoured employers and made successful litigation for tort claims a long, costly and difficult process (Epstein, 1982; Goodin and Schmidtz, 1998: 156–58). It was often difficult to establish responsibility in the face of casual complexities and epistemic difficulties in finding fault, which meant that workers were systematically disadvantaged, and employers were often exculpated and rarely had to pay out compensation. In 1884, the Workers’ Accident Insurance Act established a system of mandatory workplace insurance for employees. The statute precluded employees from bringing tort claims against employers for workplace accidents, but it ensured that employees received compensation for injuries without having to establish fault. The system became the model for modern systems of workplace insurance that exist in many mature legal systems today.

In the same way that the causal complexities in workplace accidents that make it difficult to establish fault motivated a turn to insurance-based systems of compensation, the causal complexities in establishing responsibility for climate displacement motivate a turn to an insurance model for addressing climate displacement. On this model, all of those who face displacement are owed certain duties, which vary according to the different forms that displacement can take, regardless of whether or not their displacement can be attributed to climate change. Climate change, on this model, operates as a risk multiplier, that makes incidences of displacement more likely. In the same way that employers who make their employees engage in particularly risky forms of work pay higher insurance premiums to offset the predictable costs associated with that risk, we should hold high-emitters responsible to bear a higher share of the costs associated with the increased risk of displacement associated with their responsibility for climate change. The regimes of governance that we have for addressing displacement are, on this model, akin to the systems of insurance that ‘pay out’ when individuals find themselves displaced. The costs that states are required to bear that stem from their responsibility for climate change are akin to higher insurance premiums.

The insurance approach could be institutionalised in a number of ways. For example, a Protocol to the UNFCCC could require states to contribute additional funds to the financing of displacement governance in proportion to their contributions to climate change. Those who have advocated insurance-based approaches to climate displacement so far have tended to focus on specific insurance-based policy instruments, which are one way of translating this general idea into real-world policy (Thornton, 2018: 98–127; Penz, 2010). The point that I am making here is more basic: at the level of principle, the insurance model of responsibility enables us to reconcile the complexity of climate displacement with the questions of responsibility that it raises. So long as we adopt the insurance model of responsibility, the pluralist theory can answer to the questions of responsibility that arise in the context of climate displacement.

Conclusion

This paper has sought to set an agenda for a political theory of climate displacement, by critically examining the monist approach and by proposing an alternative. The monist approach relies on the concept of the climate refugee, which obscures the complexity and heterogeneity of climate displacement. This means that it will fail to treat like cases alike, because the complexity of climate displacement means that we will be unable to reliably distinguish individual climate refugees from other displaced persons. And it will fail to treat relevantly different cases different, because the heterogeneity of climate displacement means that there are morally relevant differences between cases that it ignores.

In response to these failings, I outlined a pluralist theory of climate displacement. The pluralist theory disaggregates the different practical and institutional contexts in which climate change and displacement interact. I distinguished between five contexts of climate displacement: community relocation, territorial sovereignty, labour migration, internal displacement, and refugee movement. At the same time, climate displacement is unified by the distinctive considerations of responsibility that it raises. The pluralist theory reconciles these considerations of responsibility with the complexity of climate displacement by focusing on how responsibility for climate displacement affects the second-order question of how the burdens of tackling displacement should be shared, rather than the first-order question of what we owe to the displaced.

Here, I have only been able to sketch some key questions that a theory that a pluralist theory of climate displacement will need to address. A fully elaborated pluralist theory of climate displacement will tell us what principles and practices we should adopt in each context of climate displacement. My contention is that developing such a theory is a more promising way forward for a political theory of climate displacement than a proposal that focuses on the concept of the climate refugee.