Refugees, membership, and state system legitimacy

Abstract

In the literature on refugeehood in political theory, there has been a recent turn towards what have been called “state system legitimacy” views. These views derive an account of states’ obligations to refugees from a broader picture of the conditions for international legitimacy. This paper seeks to develop the state system legitimacy view of refugeehood by subjecting the most developed version of it—the account developed by David Owen—to critical scrutiny. We diagnose an ambiguity in Owen’s theory of refugeehood, in the concept of political membership, and unpack the implications of this ambiguity for state system legitimacy views. First, we reconstruct the key aspects of Owen’s account of refugeehood and show how it represents an advance over competing theories. Then we discuss the methodological underpinnings of Owen’s account, showing the constraints and opportunities faced by state system legitimacy views. Next, we raise some problems for the conceptual distinctions that Owen develops between different types of refugee protection: asylum, sanctuary, and refuge. The underlying feature that leads to these problems is an ambiguity in the concept of political membership, which is at the core of Owen’s view of refugeehood. Finally, we distinguish two interpretations of political membership in the institution of refugeehood and chart out some possible ways forward for state system legitimacy views. The critique developed here is a sympathetic one, aimed at the further development of state system legitimacy views.

Keywords:

• Refugees
• Membership
• International Legitimacy
• Human Rights
• Civic Status

In the literature on refugeehood in political theory, there has been a recent turn towards what have been called ‘state system legitimacy’ views (Sharp Forthcoming), such as those developed by David Owen (2020), Gillian Brock (2020) and Chris Bertram (2018; see also Carens 2013, 196). These views derive an account of states’ obligations to refugees from a broader picture of the conditions for the legitimacy of the international state system. This turn represents a promising and novel line of inquiry, but it remains under-theorized. This paper seeks to develop the state system legitimacy view of refugeehood by subjecting the most developed version of it – the account developed by David Owen (2020, 2019, 2016) – to critical scrutiny.¹ We diagnose an ambiguity in Owen’s theory of refugeehood, in the concept of political membership, and unpack the implications of this ambiguity for state system legitimacy views. In doing so, we build on earlier work that takes the concept of political membership to be of central importance in the normative foundations of the refugee regime (Buxton 2021).² The critique developed here is a sympathetic one, aimed at the further development of state system legitimacy views.

We take Owen’s theory of refugeehood as an exemplar here for several reasons. First, Owen’s defence of his theory shows how state system legitimacy views represent a theoretical advance over competing views of refugeehood in political theory. Second, an examination of the methodological underpinnings of Owen’s account reveals the constraints and opportunities faced by the most promising versions of the state system legitimacy view. Third, the tripartite distinction between asylum, sanctuary, and refuge that provides the conceptual core of Owen’s reveals the ambiguity over the concept of political membership that forms the centre of our critique. Finally, we take Owen’s account of the institution of refugeehood to be the strongest version of the state system legitimacy view of refugeehood in the literature. If we can show that there is an important ambiguity within his account, this will bear on the prospects of state system legitimacy views more generally.

The paper proceeds as follows. First, we reconstruct the key aspects of Owen’s account of refugeehood and show how it represents an advance over competing theories. Then we discuss the methodological underpinnings of Owen’s account, showing the constraints and opportunities faced by state system legitimacy views. Next, we raise some problems for the conceptual distinctions that Owen develops between different types of refugee protection: asylum, sanctuary, and refuge. The underlying feature that leads to these problems is an ambiguity in the concept of political membership, which is at the core of Owen’s view of refugeehood. Finally, we distinguish two interpretations of political membership in the institution of refugeehood and chart out some possible ways forward for state system legitimacy views.

Reconstructing refugeehood

Owen’s account begins with a diagnostic project: one of identifying and tracing the history of two ‘pictures’ of refugees (4–11). The humanitarian and political pictures represent two broad positions on the question of ‘who counts’ as a refugee, and of two currents in the practices of refugee protection.

The humanitarian picture locates the claims of refugees as falling under the more general ‘moral duty to prevent undeserved suffering’ (7). This view identifies refugees as those who face threats to their basic needs, whatever the source or character of those threats. Owen identifies Betts and Collier (2017) as contemporary exponents of the humanitarian picture (5), and others such as David Miller (2016) and Chandran Kukathas (2016) adopt a similar perspective. The humanitarian picture is associated with the United Nations High Commissioner for Refugees (UNHCR)’s self-understanding as a provider of humanitarian assistance and with the de facto and de jure expansions of its mandate since the emergence of the modern refugee regime (27–29).

The political picture bases the claims of refugees on ‘the underlying political obligation to redress the injustice of membership repudiation to which refugees are subject’ (7). Rather than the fulfilment of basic needs, refugees are owed asylum as a surrogate form of political membership (7). Owen identifies Max Cherem (2016) as a contemporary exponent of the political picture (5), and others such as Matthew Price (2009) adopt a similar perspective. The political picture is associated with the 1951 Refugee Convention’s focus on persecution (26–27).

Owen’s task in distinguishing these pictures is genealogical: he seeks to trace the history of our conceptions of refugeehood and to show how they have come to hold sway over us, in doing so freeing us from our inherited understandings of refugeehood. His genealogical analysis locates the points of agreement and disagreement around which most political theorists working on refugeehood have coalesced and explains why competing positions have become ossified: theorists have been talking past each other, because they have in mind different visions of the point and purpose of the refugee regime. Whereas those operating within the ‘humanitarian’ picture see the point of the refugee regime as addressing the plight of those who fall below some threshold level of basic needs, those working from the ‘political’ picture see it as addressing a specific form of membership-related harm.

After distinguishing these two pictures, Owen offers a ‘normative reconstruction’ of the institution of refugeehood that seeks to move past the impasse between these views (34). This view begins with a view of the international order as a regime of governance – produced and maintained through the actions of states that share a collective responsibility for its functioning – which is structured by the norms of state sovereignty/non-intervention and human rights. The international order’s claim to legitimacy depends on being able to reconcile these two commitments. The most basic way that this is done is through the assigning of people to states, on the basis of the norm that states are responsible for protecting their own citizens’ human rights and the presumption that they are willing and able to do so (44–46). When states fail to protect their own citizens, this poses a legitimacy deficit for the international order. The general function of the institution of refugeehood, on Owen’s view, is to repair the legitimacy of the international order along this dimension of failure. The refugee occupies a particular place in the international order as one whose state is failing, in one way or another, to provide protection of her basic rights. This lack of protection grounds her claim that the international order of states stands in loco civitatis (in the place of her state) in providing this protection. Refugees are:

persons whose basic rights are unprotected by their state and can be protected only through recourse to the international order of states acting in loco civitatis and thereby affirming the dual commitment of the regime of global governance to the norm of state sovereignty and human rights (50).

The primary duty that the international order owes to all refugees is the duty of non-refoulement – that is, the duty not to return refugees to a situation in which their basic rights are under threat (51). Beyond this, however, Owen argues that we should distinguish between different kinds of refugees with different normative claims. The specific content of refugee protection will depend on the particular normative status that the refugee has, which corresponds to three differing standards of protection: asylum, sanctuary, and refuge.

In the case of asylum, the refugee faces persecution by the state (or is wilfully unprotected by the state from private persecution). Persecution is a distinctive wrong, which consists in ‘the denial of one’s equal standing as a member of the state.’ In the context of an international order divided into sovereign states, amounts to the denial ‘of one’s equal standing as a member of global political society’ (55). The asylum refugee is in a precarious condition: without this standing, one is ‘acutely vulnerable both to the contingencies of circumstance and to the agency of public or private others’ (55.) The warranted response to the condition of the asylum refugee is the grant of surrogate membership in another state, which reasserts the entitlement to political standing in the international order (56).

In the case of sanctuary, the refugee is fleeing generalized violence or breakdowns of public order. Such an individual’s political membership becomes ‘harmfully ineffective’ (58), but it has not been wrongfully denied in the same way as in cases of persecution. The sanctuary refugee is deprived of her ‘social world’ and in a position of ‘social and civic limbo’ (59–60). Sanctuary refugees ‘lack an ability that is taken for granted by citizens who conduct their lives against the background of a right to secure residence in a state, namely the ability to plan their future’ (60). The warranted response to the sanctuary refugee’s condition is the provision of access to ‘the basic security, liberty and welfare that the protective and enabling functions of citizenship would normally provide’ (58–59). This does not justify the immediate granting of new membership in another state or immunity from non-voluntary repatriation, as in the case of asylum, but there are ‘strong grounds to think that more rapid access to membership would not only be justified but also called for’ (61).

In the case of refuge, the refugee is fleeing discrete events such as famines or natural disasters that threaten her basic rights, and is so situated that she can best seek protection by crossing an international border (61–62). Those with claims to refuge must have their basic rights protected and the general obligation of non-refoulement still applies. But the status of refuge itself generates ‘no basis for claims to membership in the state’ (62). Those with a claim to refuge may eventually have an entitlement to political membership should their displacement become protracted, but that claim to membership would stem from ‘social membership constraints’ (84) – from their having become de facto social members in their state of residence (Carens 2013, 158–69) – rather than from their claim to refuge itself.

Taken together, these three categories of refugeehood, along with the more general conception of the function of the refugee regime, articulate a distinctive conception of refugeehood. Owen’s account draws from the humanitarian and political pictures, but his account moves beyond them by providing a more comprehensive view.

Pictures and practices of refugeehood

In his diagnostic project, Owen provides a genealogical analysis of the historical development of two pictures of refugeehood. In his positive project, Owen seeks to normatively reconstruct ‘the point and purpose of the institution of refugeehood’ that makes sense of this history (44). Methodologically, this aligns well with the ‘practice-dependent’ turn in political theory.³ A practice-based theory of refugeehood aims to reconcile different aspects of the history and practice of refugee protection and to fit within the normative commitments of the international order.

This practice-based methodology is a promising way of developing the state-system legitimacy view, and others like Gillian Brock (2020, 33–63) also draw on a practice-based view of the legitimacy of the international state system to derive an account of duties to refugees. Although there may not be a strictly necessary connection between the state-system legitimacy view and the practice-based methodology, there is certainly an elective affinity between them. The state-system legitimacy view’s account of refugeehood is tied to a broader account of the legitimacy of the international state system, and so a constructive interpretation of the constitutive norms of the international state system provides a natural place to start in developing such a view. A theory based on the normative standards implicit in the best interpretation of the international state system clarifies our existing practices and has justificatory purchase because it draws out the significance of our pre-existing normative commitments.

To illustrate the constraints and opportunities facing practice-based versions of the state system legitimacy view, it is worth examining two methodological questions arising from Owen’s account. The first concerns the consistency between Owen’s genealogical project and his constructive project. His genealogical project reveals the historical contingency of our practices of refugee protection through an analysis of their origins and development. The positive project constructs an account of refugeehood from those very same practices. A question naturally emerges: why should we take those practices to be firm ground from which to construct a normative theory of refugeehood, given their contingent histories?

A fuller understanding of Owen’s genealogical project supplies an answer. Often, we take genealogical explanations to be debunking. The revelation that our concepts or beliefs are artefacts of a particular history is taken to induce a sense of ‘genealogical anxiety’ – a concern that the contingency of our concepts means that they are unreliable, and that as such they should be abandoned (Srinivasan 2019). But this is too quick: the mere fact that our concepts have contingent histories does not mean that they should be discarded. Instead, genealogical criticism can reveal how our concepts and categories have structured our practices to date – the forms of life they have made possible – and illuminates how transforming them might enable us to transform our practices in a project of ‘worldmaking’ going forward (Srinivasan 2019, 140–47).

The aim of Owen’s genealogical investigation is not to debunk our conceptions of refugeehood. Rather, Owen’s project reveals that each of these pictures of refugeehood – the humanitarian and the political – are precisely that: pictures, which are only ever partial and incomplete. They are ‘pictures’ in the Wittgensteinian sense: they have ‘held us captive’ and have precluded us from considering alternative possibilities (Wittgenstein 2007, 53 [§115]; see also Owen 2002, 2003). Rather than putting us in a state of false consciousness, these pictures have held us in a state of ‘aspectival captivity’ or ‘restricted consciousness’: it is not that either of these pictures is false, but that each only shows a partial perspective (Owen 2002, 217–19). The genealogical analysis of these pictures of refugeehood is aimed at liberating us from this situation of aspectival captivity, so that we can transform our account into one which can capture the full range of our commitments. Understanding Owen’s genealogical critique helps to defend his conception of refugeehood against the potential methodological objection that his account is built from the same conceptions that his genealogical analysis gives us reason to reject.

A second methodological question, however, remains open. Indeed, it remains necessarily open on the practice-based method. Owen’s theory of refugeehood is constructed from, and seeks to make sense of, the practices of refugee protection. But any practice-based theory will inevitably foreground some aspects of the practice and background others. One way to challenge Owen’s theory is in terms of the aspects of the practices of refugee protection that it makes most and least salient. A practice-based normative theory requires us to engage in an interpretation of our practices. The aim of an interpretation is, in Charles Taylor’s (1971, 3, 5) words, to ‘bring to light an underlying coherence or sense’ which is present in only ‘a confused, fragmentary, cloudy form’ in our practices. But our political practices are messy, complex and – crucially – subject to disagreement and contestation. In addition to the official aims and rules that structure the practice, the ‘operative’ concepts in a practice – those that are ‘hidden’ or ‘implicit’ (Haslanger 2012, 370) – will also bear on how we should understand its function. When we reconstruct a concept such as refugeehood, we make choices about which elements of the practice to foreground or include, which will have a significant bearing on the normative reconstruction that we develop.

This need not be problem for a practice-based approach to refugeehood. It simply means that we should not see the process of reconstruction as one that uncovers a fixed or immutable truth about the nature of the concept of the refugee that was there all along. Rather, as Robert Brandom (2002, 14) argues, this mode of reconstruction consists in ‘offering a selective, cumulative, expressively progressive genealogy’ of the concept in question, through which we ‘make’ a reconstruction of a concept: ‘it is insofar as one takes the tradition to be rational […] that one makes the tradition be and have been rational.’ The commitments that we bring to bear on this inquiry inevitably shape our interpretation, meaning that the philosophical theory and the interpretation of the practices that we develop are ‘two sides of the same coin, two aspects of one enterprise’ (Brandom 2002, 16). For Brandom, the upshot of this is that not that reconstruction is misguided. Rather, it is that the sense in which a reconstruction can be ‘correct’ is ‘not an exclusive one.’ Reconstruction of this kind leaves open space for alternative interpretations, and the ‘best philosophical response to such a narrative is not belief or endorsement, but rather the telling of more such stories’ (Brandom 2002, 16).

In this spirit, we can examine Owen’s reconstruction of the institution of refugeehood in terms of one aspect of the practice it foregrounds. Owen’s reconstruction makes the expressive function of the refugee regime particularly salient. On Owen’s view, the refugee regime expresses a commitment to the constitutive norms of the international order: those of human rights and state sovereignty, and each specific form of protection – asylum, sanctuary, and refuge – expressive more specific commitments (63–64). The grant of asylum, on this view, expresses condemnation of states that persecute their citizens or fail to protect them from private persecution (55). This claim accommodates an argument developed by Price (2009, 70), who argues that in providing asylum, a state ‘directs condemnation toward other states for having egregiously mistreated the refugees it protects.’ Owen’s interpretation accommodates this view of the expressive function of refugeehood, but unlike Price, he does not use it to generate a restrictive account of ‘who counts’ as a refugee.

Owen makes this aspect of the refugee regime particularly salient. But there are alternative readings of the expressive function that give us reason to be reticent about giving it a central place in a normative theory of refugeehood. One such alternative reading, set out by T. Alexander Aleinikoff (1992), is that the provision of asylum served the function of enabling Western states to ‘score ideological points’ against communist regimes during the Cold War by accepting refugees for permanent resettlement. This explanation accounts for ‘exilic bias’ of the refugee regime during that era. But more recently, this ideological function has receded in importance. Aleinikoff suggests that this explains why that we have witnessed a shift to ‘state-centred’ forms of refugee law, which prioritize repatriation over asylum, as states seek to contain refugee flows within regions of origin (Aleinikoff 1992, 125–30).

This interpretation of the practices of refugee protection creates some unanswered questions for Owen’s account. If the expressive nature of asylum were really about scoring ideological points, then would we want this to feature in our normative reconstruction of refugee protection? Wouldn’t this make the refugee regime is an instrument of ideological conflict? And if, as Aleinikoff suggests, this aspect of refugee protection has become less significant in practice in recent history, should its importance likewise recede in our normative theory of refugeehood, or is something that we should seek to recover?

To be clear, the point of this critique is not to challenge Owen’s method in reconstructing the institution of refugeehood. It is rather to bring to light the important point that a practice-based theory of refugeehood does not inevitably require us to adopt his reading of the history and practices of refugee protection. If we think that the practice-based method is a promising way of developing the state system legitimacy view, then it is important to be cognizant of both the constraints and the opportunities involved in this project. A practice-based theory of refugeehood should be developed from an interpretation of the history and practices of refugee protection. But this does not mean that it needs to adopt the substantive commitments displayed by Owen’s account wholesale. We return to this implication of the practice-based methodological approach for developing the state-system legitimacy view in Section 4.

Asylum, sanctuary, and refuge

A central feature of Owen’s account is the claim that the normative status of individual refugees should be ‘differentiated’ according to the three categories of asylum, sanctuary, and refuge. In this section, we apply some pressure to the distinctions between these categories. We raise a range of objections to Owen’s tripartite distinction as it is presently formulated here, but our aim is not to demonstrate conclusively that this distinction should be rejected. It is rather to show that – as we discuss in Section 4 – the common feature that underlies these concerns is an ambiguity in the concept of political membership.

Asylum/sanctuary

One key aspect of the distinction between asylum and sanctuary is the status of the refugee’s political membership in their state of origin. In the case of the asylum refugee, their status as a member is wrongfully denied whereas in the case of the sanctuary refugee, their status as a member is merely ‘ineffective’ or ‘inoperative’ (58). For Owen, the asylum refugee has a claim not only to the protection of their basic rights, but also to be granted new membership in a host state, as a way of reaffirming their political standing in international society (56). The sanctuary refugee does not have the same claim to membership in a host state. Whilst they may have a claim to expedited access to membership on the basis of a situation of ‘social and civic limbo’ they are in, sanctuary ‘does not require rapid admission to a new membership’ (59–60).

There are at least three possible reasons why the denial of political membership could warrant the immediate provision of new membership in a host state: (i) providing new membership serves the expressive function of condemning outlaw states; (ii) the provision of new membership is distinctively suited to address the moral wrong involved in the denial of membership; (iii) the provision of new membership responds to situations where individuals do not have (effective) political membership. In our view, (iii) is the most plausible justification, but accepting (iii) makes it difficult to maintain the distinction between asylum and sanctuary.

The alternative reading of the expressive function of asylum adduced above provides us with some reason to be sceptical of claim (i). But beyond this, there is a more basic reason that we can reject the role of expression in grounding the distinctive nature of asylum: appealing to the expressive function of the refugee regime to justify the provision of new membership appeals to the wrong kind of reasons. Whether a refugee has a claim to new political membership should depend on her own distinctive needs or circumstances, rather than on the expressive function of granting asylum in international politics. To see this, imagine a case where the expressive function of the refugee regime has no effect – where a state is so irredeemably repressive that expressive condemnation of its persecution has no prospect for changing its behaviour. In such a case, the expressive condemnation of that state by the international community is pointless, and so it is unclear why those fleeing from that state would have a claim to political membership elsewhere – after all, this would not serve the broader expressive function of the refugee regime. But surely, we want to say those fleeing from this state are precisely those with the strongest claim to new political membership. There may be other reasons to express condemnation – for example, such condemnation might signal to other, less irredeemable, states that certain behaviours are wrong, or might be non-instrumentally valuable in itself – but the broader point is that these kinds of reasons are not a firm basis for an account of refugeehood. If we base who counts as a refugee principally on whether and how condemnation is expressed, then we are no longer appealing to the position of the refugee at all. In other words, an account that places too much emphasis on expression ‘looks like it is nothing to do with the position of the refugee herself’ (Buxton 2021, 6).

To be clear, we do not think that this expression-first approach is the best interpretation of Owen’s argument – although he seeks to accommodate the expressive function of the refugee regime, his differentiation between refugees is also designed to respond to the distinctive normative claims that refugees themselves raise (53–54). But it is nonetheless useful to clarify why this is not a good reason for the asymmetry in the rights of those with claims to asylum and sanctuary.

Claim (ii) is more promising and a better interpretation of Owen’s view. On Owen’s view, the international community has ‘an obligation to assert, through the granting of new membership, the political standing of persons who have been wrongfully denied it by their home state’ (56). The wrong of persecution leads to a serious harm, which is that ‘by being rendered de facto stateless, one is made acutely vulnerable both to contingencies of circumstance and to the agency of public or private others’ (55). On this view, persecution is a distinctive wrong that leads to a serious harm: the harm of being de facto stateless. What should we make of this view?

First, it is important to recognize that persecution is not the only way in which this harm can be brought about – as Owen recognizes – since refugees can be rendered de facto stateless in other ways, such as through state collapse.⁴ Consider the two million or so refugees who fled Zimbabwe to South Africa between 2000 and 2010. Although some faced individualized persecution under Mugabe’s regime, many had simply fled generalized conditions of severe deprivation and state failure (Betts 2013, 54–77). In this case, those affected had lost the protections associated with membership. They had no state to which they could appeal for the protection of their basic rights and were vulnerable to the contingencies of circumstance and the agency of others. But this was also clearly not a case of persecution at the hands of state authorities (or permitted by state authorities).⁵ So, the explanation here cannot be that the wrong of persecution grounds a claim to new membership because it uniquely involves the harm of de facto statelessness.

Instead, the most plausible interpretation of claim (ii) is that the grant of new membership responds to the wrong of persecution itself, rather than the harm that persecution brings about. On this interpretation, it is only when the harm of de facto statelessness is accompanied by the wrong of persecution that it triggers a claim to asylum, understood as involving a claim to new membership in a host state. The grant of new membership, on this view, is a restitutive obligation that responds to the distinctive wrong of persecution (71).

The basic problem with this argument is that the logic of the state system legitimacy view suggests that the refugee regime should respond to the harm of de facto statelessness rather than the wrong of persecution. On the state system legitimacy view, the basic function of the refugee regime is to ensure that those whose states fail to comply with the norm of protecting their citizens’ basic rights have firm guarantees of protection from the international community. If a refugee’s basic rights are insecure when they lack effective political membership, then the provision of new political membership appears obligatory regardless of whether they are in this position because of their state’s lack of will or lack of capacity. It is the harm itself that generates the legitimacy problem for the international order, rather than the wrong and harm taken together.

One final way in which one might try and defend claim (ii) would be to maintain that there is a difference in the harms that arise in each case. One area in which this could come to light is through a consideration of rights of external citizenship, such as consular support abroad. Refugees displaced by generalized breakdown of the state – claimants of sanctuary – may still enjoy support from their state when abroad. For instance, many refugees displaced by civil war in Latin America still receive diplomatic support when living in America; they still have the right to vote externally and can seek assistance from their relevant embassy. But refugees who are actively persecuted by their state cannot access these citizenship rights. As such, the external as well as internal elements of citizenship are closed off to them. At first glance, this appears to locate an important difference between persecuted and non-persecuted refugees.

One issue with appealing to this difference is that it does not track the asylum/sanctuary distinction perfectly: many states facing generalized breakdown also lack the ability to protect their citizens when abroad. This might obviously be the case in examples of ‘failed states’ where some countries lack even the ability to offer consular support. So, even though persecution may more commonly result in a loss of external citizenship, those who qualify under Owen’s framework as sanctuary refugees can face this harm as well.

The more basic problem with appealing to this distinction, however, is that it does not mark out a difference that is morally relevant when it comes to the ascription of new membership. Even if sanctuary refugees have access to the rights of external citizenship, this does not suffice to guarantee them the robust forms of rights protection associated with political membership that they lack. Without the ability to provide protection on their territory – which ex hypothesi they lack – states that produce sanctuary refugees cannot provide firm guarantees of rights protection merely through diplomatic and consular protection. Even when sanctuary refugees have access to the rights of external citizenship, they are still in a situation of vulnerability to the contingencies of circumstance and the agency of others. So, even if there is a difference in the harms suffered by sanctuary and asylum refugees in terms of external citizenship, that difference does not justify the asymmetry between them in terms of access to new political membership.

To be clear, we think that Owen’s view has the resources to account for these cases. But in order to do so, he should appeal directly to claim (iii), that there is a universal entitlement to political membership somewhere and that the provision of new membership responds to situations where individuals do not enjoy such membership. Owen claims that ‘[i]n a world in which state membership is the basic condition of political standing, it is a duty of the international order of states to ensure that all persons enjoy such standing’ (56). This claim is at the core of the state system legitimacy view, and could well be glossed as an interpretation of Arendt’s (1951, 297) claim that there is a universal ‘right have rights.’ But contra Owen, this claim suggests that whoever does not have effective political membership has an entitlement to have such a status reinstated – whether they lack this status because of persecution or otherwise. Otherwise, the international order is failing in its duty to ensure that all persons enjoy political standing in some state. Those whose membership is ‘ineffective’ or ‘inoperative’ appear to have an equal claim to new political membership on these grounds. This makes the conceptual distinction between asylum and sanctuary difficult to maintain.

Sanctuary/refuge

Next, we turn to the distinction between sanctuary and refuge. Sanctuary is owed to those who are fleeing conditions of generalized violence and insecurity where the state has lost the capacity to protect its members’ basic rights. Refuge, by contrast, is owed to those who are fleeing discrete and specific events, such as natural disasters and famines, ‘where a person is so situated that she can save herself from the threat to her basic rights posed by the event in question by seeking immediate shelter across an international border and this is her best reasonable option in the circumstances in which she finds herself’ (62).

There is also a distinction between sanctuary and refuge in terms of access to political membership. Whilst sanctuary ‘does not require rapid admission to a new membership,’ Owen does argue that there are nonetheless good grounds to grant sanctuary refugees expedited access to political membership – albeit not on the same basis as those owed asylum. Owen argues that expedited access to political membership can be justified for sanctuary refugees because they are ‘deprived of [their] social world,’ put in a situation of ‘civic limbo,’ and because they face a ‘distinctive challenge’ in that they lack effective citizenship (56–60). Refuge, by contrast, entitles its claimants to immediate humanitarian assistance, but their status generates ‘no basis for claims of membership in the host state,’ except in cases where displacement becomes protracted (62). When claimants of refuge do have a claim to political membership, that claim is based in not based not on the status of refuge, but rather on the basis of ‘social membership constraints’ (84): like lawfully resident migrants, their claims to membership become stronger the longer they are within the host state, since they establish social ties and become de facto social members.

The claim that sanctuary refugees face a ‘distinctive challenge’ resulting from their lack of effective citizenship could be interpreted in two ways. On the first reading, the distinctive challenge faced by sanctuary refugees is that they are ‘deprived of their social world’ and put in a situation of ‘civic limbo.’ These claims arise in situations of sanctuary because, according to Owen, it is a ‘pervasive feature of refugee crises that their time horizon is liable to be indeterminate’ (59). The social dislocation of refugees is especially troubling, and the situation of civic limbo is particularly acute, because there is often no clear and foreseeable end to the predicament of sanctuary refugees. Thus, they are entitled to expedited access to political membership in order to remedy this predicament.

The problem with this justification is that these features of sanctuary will also apply to many claimants of refuge. The time horizon of a particular refugee crisis will not always map onto the distinction between sanctuary and refuge. Discrete events such as disasters may well – and often do – result in protracted situations of displacement (McAdam 2020, 585). The reverse is also true: situations of generalized insecurity, or even persecution by oppressive regimes, may be relatively short-lived. Gibney (2015, 454) points out that oppressive regimes often produce refugees ‘in their dying days’ whereas other basic rights threats, such as chronic poverty, can ‘drag on for many years.’ Something similar may also be true of situations of generalized insecurity and violence – refugees may be produced in the short period before order and the rule of law become firmly established. So, it does not seem that these claims mark out a significant difference between sanctuary and refuge that would justify an asymmetry in access to political membership.

On the second reading, however, there is a more fundamental difference in terms of the relationship between the refugee and her home state. In situations where sanctuary is appropriate, refugees lack ‘the basic security, liberty and welfare that the protective and enabling functions of citizenship would normally provide’ (58–59). In situations where refuge is appropriate, refugees face threats to their basic security, liberty, and welfare which can be best averted by crossing a border, but it is not clear that their citizenship status in the home state is ‘ineffective’ or ‘inoperative’ in the same way that it is for sanctuary refugees. One way of interpreting this difference is as the difference between cases where states have lost the standing to act as the guarantor of their members’ basic rights (sanctuary), and cases where there are particular human rights deficits, but those deficits do not meet the threshold that would need to be met for the state to lose its standing to act as the guarantor of its members’ human rights (refuge).⁶

This interpretation of the sanctuary/refuge distinction makes sense of Owen’s claim that membership is ‘ineffective’ in situations of sanctuary. But it also presents a dilemma for Owen’s view: either those with a claim to refuge are refugees proper, in which case their membership must be ineffective, or they are not best understood as refugees at all. On the state system legitimacy view, refugees are persons whose basic rights are unprotected and for whom the international community must stand in loco civitatis – in other words, for the international community must act so as to replace, rather than merely supplement, the home state (48). If those owed refuge are refugees in this sense, then this must be because they not only face particular deficits in their basic rights, but also because their relationship with their home state has broken down in such a way that makes it appropriate for the international community to stand in loco civitatis. But if those owed refuge still stand in the right kind of relationship with their state for it to act as the ongoing guarantor of their human rights, then the international community does not need to stand in loco civitatis – its proper role is to supplement, not replace, the home state. Thus, on this interpretation of the distinction between sanctuary and refuge, those owed refuge are, properly speaking, not refugees at all.

This need not mean that the international community has no duties to such persons. The international community may well still have duties of protection. But such duties need not be understood in terms of the refugee regime’s specific function, which is to maintain the legitimacy of the international order by requiring the international community to replace the home state as the guarantor of the basic rights of its members. Such protection might be better viewed as belonging to another normative framework, such as the framework of ‘complementary protection’ (cf. Lister 2019). Owen argues that refuge ‘serves the same basic function as emergency assistance’ within a state and is ‘essentially part of the same emergency assistance policy toolkit’ (62). He argues that the international assistance regime also constitutes a legitimacy-repair mechanism for the predictable failures of the international order (47) – but crucially, it has a distinct function to the institution of refugeehood. On this interpretation of the sanctuary/refuge distinction, which allows for a more fundamental difference between the two categories, those with a claim to refuge are not properly viewed as refugees at all.

Perhaps the distinction between sanctuary and refuge could be maintained if we take the norm of non-refoulement to be morally basic in distinguishing refugees from other claimants of emergency assistance. Those with a claim to refuge might be distinguishable from others with a claim to emergency assistance on the basis that they, like all other refugees, are protected by the rule of non-refoulement. This response is suggested by Owen’s claim that the norm of non-refoulement is central to understanding the institution of refugeehood (31–32). One problem with this response is that the norm of non-refoulement can itself be interpreted in two ways, each of which poses problems for Owen’s account. If it applies to only those outside of the territory of their state, then Owen’s account can no longer count those in ‘safe havens’ within their states as refugees. For this reason, Owen argues for a more capacious interpretation of non-refoulement that applies within the state as well, understanding it as including an obligation not to return refugees from safe havens (32). But if we adopt this more capacious interpretation of non-refoulement, then it is not clear that it can do the work that it is supposed to do to distinguish claimants of refuge from claimants of other forms of emergency assistance. After all, others with claims against being returned – such as internally displaced persons – are protected by non-refoulement understood in this broad sense, but are not refugees.

If we are to maintain the claim that those who are owed refuge are refugees proper, then we need to interpret the condition of those who need refuge as one in which their membership is ineffective or inoperative in some way. But if this is true, then there no longer seems to be a meaningful distinction between those who need refuge and those who need sanctuary. Sanctuary refugees alone no longer seem to face a ‘distinctive challenge’ of lack of effective citizenship; those with a claim to refuge face such a challenge as well. And if this is true, then it appears that those who are owed refuge have the same claim to expedited access to political membership as those who are owed sanctuary.

Two faces of membership

Although we have raised several distinct objections to Owen’s tripartite distinction, they share a common root: an ambiguity in the concept of political membership. When viewed from the perspective of membership, each category of refugeehood appears to collapse into another. Having one’s membership repudiated (as in the case of asylum) seems to place one in the same position of vulnerability that one faces where one’s membership is merely ineffective (as in the case of sanctuary). Where one’s human rights merely go unfulfilled (as in the case of refuge), one’s membership either appears to be ineffective (as in the case of sanctuary), or one appears not to count as a refugee at all.

The basic idea that the concept of membership is central to the institution of refugeehood is a key insight of the state system legitimacy view (Buxton 2021). In order to develop state system legitimacy theories of refugeehood, then, it is important to build a unified theory of membership and its relation to refugee status. In this final section, we map out a way forward for state system legitimacy theories like Owen’s, by distinguishing between two possible interpretations of membership that could figure in such a theory: a human rights interpretation and a civic status interpretation.

Importantly, these two interpretations need not be viewed as being mutually exclusive. Civic status may be necessary for the protection of human rights, and the right to a formal civic status – the right to nationality – is an internationally recognized human right. What we claim here instead, then, is that human rights and civic status function as two ‘faces’ of membership that might be foregrounded in different theories of refugeehood. It is plausible to suggest that Owen’s account of the institution of refugeehood, for example, is one which captures both faces. The criticism here is not with that both faces are present in the theory of refugeehood. But we do claim that an ambiguity around which kind of membership is central to the state system legitimacy view leads to a confusion about the standards for international legitimacy itself. If Owen is appealing to both accounts – civic status and the protection of human rights – then we need to know what role each face of membership plays when it comes to the standards of legitimacy for the international order, since each face will have different implications. In other words, the face of membership we put at the centre of a theory of refugeehood makes an important difference to the broader shape of that theory.

Before clarifying these features of membership, it is notable that the interpretations of membership set out here are interpretations for the purposes of a theory of refugeehood. They are not intended as comprehensive accounts of the concept of membership for all contexts. Rather, they are meant as ‘ameliorative’ (Haslanger 2012) interpretations of the concept of membership as it figures within a system state legitimacy theory of the institution of refugeehood. The basic premise of state system legitimacy theories of refugeehood is that the international community has a collective duty of legitimacy to protect the status of all persons as members, and that the institution of refugeehood is a way of providing surrogate membership to those whose membership status is in question. The question that we attempt to address here is this: what does it mean for an individual to be a member in the sense that is relevant to the institution of refugeehood?

On the human rights interpretation, membership of a state is viewed primarily as an instrument for the protection of basic rights. Ordinarily, it is through one’s status as a member of a particular political community that one enjoys rights protections. According to the of norms the international order, each state bears primary responsibility for the protection of the human rights of its own members (Beitz 2009). Citizenship is a formalized ‘international filing system’ for distributing the duties to protect basic human rights among states in the international order (Brubaker 1992, 31). On this interpretation, the point of the institution of refugeehood is fundamentally to ensure that human rights remain universally protected when states fail to fulfil their duties to their own members. Arguably, this is the view that underpins James Hathaway’s (1991) interpretation of international refugee law, given that he views refugee law as ‘continuous’ with the human rights regime. The aspect of membership that we should care about for the purposes of the institution of refugeehood, on this view, is simply its function in protecting human rights. Therefore, the moral demand that hosting refugees makes on states is to protect the rights of those whose states have failed to protect them. If states have further duties beyond protecting the basic rights of refugees that they host, such as duties to facilitate naturalization, then those duties are either only instrumentally justified as a means of ensuring that basic human rights remain universally protected or are justified by reasons that are external to the normative logic of the refugee regime itself.

On the civic status interpretation, membership is not viewed merely as an instrument for protecting basic rights. Rather, membership is a civic status that constitutes individuals as agents with political standing in their states and, through their states, in international society. This interpretation of membership is a more demanding one, which views members not merely as beneficiaries of state protection, but as active participants in the construction of the shared social world. The denial of one’s civic status as a member of a political community denies one’s ability to act in public and participate in the construction of the shared social world. There are a number of ways in which this idea could be rendered, but it is in the tradition of civic republicanism that this face of membership has played a central role. In the context of refugeehood, one important contribution that can be understood in these terms has been made by Felix Bender (2021, 2022). Bender argues that refugeehood is a response to the denial of one’s public autonomy vis-à-vis the members of one’s own state. To this, we would add that insofar as one’s civic status within one’s state is denied, one is also denied civic status within international society. In a world divided into sovereign states, it is through membership in a state that individuals come to participate in the construction of a shared international order. The function of the institution of refugeehood is, on this interpretation of membership, to reassert the civic status of those who lack membership as members of international society, by providing them with a status in a surrogate state. This interpretation more easily accounts for expansive duties, such as duties of naturalization, to be owed to refugees as a matter of satisfying the internal normative logic of the refugee regime itself (Buxton 2021).

To be clear, proponents of the civic status interpretation need not deny that membership is instrumentally valuable for guaranteeing rights in the international state system. Those who enjoy membership in a state – both formal and effective – will enjoy a more stable form of rights protection. Rebecca Buxton (2021) has recently argued that one reason that we states should naturalize refugees is because robust forms of political membership ensure the ‘claimability’ of their basic rights. The promotion of political standing can thus bolster rights protection more generally. Those who can act in public and participate in the construction of a shared social world may well be acting to secure their human rights and the rights of others. Thus, an instrumental argument for naturalization is consistent with the civic status view. The claim of the civic status interpretation, however, is that membership in the state is not only about the protection of rights. Instead, membership has value beyond mere human rights protection.

As we’ve already noted, Owen’s theory of refugeehood is perhaps best interpreted as embodying a commitment to both interpretations of the value of membership. This interpretation may help to explain – and possibly to alleviate – some of the pressure that we have applied to his ‘differentiated’ view above. Perhaps the differences between asylum, sanctuary, and refuge can be vindicated through an appeal to a hybrid theory of membership that emphasizes these two faces – we leave that as an open question here.

But even if they can be, there remains an explanatory gap in state system legitimacy views. If the refugee regime is, as Owen suggests, a legitimacy-repair mechanism, these two interpretations of membership supply different answers to the question: under what conditions is the international system of states legitimate? On the human rights interpretation, the state system is legitimate insofar as it guarantees the protection of the basic rights of all persons. On the civic status interpretation, the state system is legitimate insofar as it guarantees the civic status of all persons as agents with political standing in international society. Both ways of foregrounding the value of membership are coherent on their own terms. To adjudicate between them, we must ask which interpretation supplies a more plausible understanding of the normative demands of the legitimacy of the international order. To combine them, we need a more complex account of the conditions under which the international order is legitimate.

One way of moving forward here would be to return to the methodological underpinnings of Owen’s articulation of the state system legitimacy view. As we have seen, Owen’s articulation of the state system legitimacy view is developed from a normative reconstruction of the practices of refugee protection in the international order. On this methodological approach, one way to judge the proper place of either the civic status or the human rights ‘faces’ of membership is to consider which provides the better interpretation of the place of membership in a reconstruction of the international order and the conditions for its legitimacy. Insofar as proponents of the state system legitimacy view adopt this methodological approach, then, they will face some constraints. But we have also seen that such a reconstruction may be more flexible than we might initially think. There are different currents and trends in the practices of the international order, and which of those currents and trends figure in our interpretation of the various practices will partly be a matter of the commitments that we bring to bear on the inquiry. Thus, there are also opportunities for more radical interpretations of the demands of the legitimacy of the international order.

Overall, our analysis suggests that competing conceptions of the refugee regime are likely to depend on background disagreements about the conditions for the legitimacy of the international order as a regime of governance. Bringing to light these more fundamental questions may help to navigate a way forward for state system legitimacy theories.

Conclusion

In this paper, we have sought to develop state system legitimacy views of refugeehood by critically examining David Owen’s theory of refugeehood. The aim of our analysis has been to use Owen’s theory as a lens through which the challenges facing a state system legitimacy view of refugeehood can be exposed. The principal difficulty uncovered here is an ambiguity in the concept of political membership, which we illustrated through a critical engagement with Owen’s tripartite distinction between asylum, sanctuary, and refuge. We suggested that there are two faces of membership that could be foregrounded within a state system legitimacy view: a civic status and a human rights interpretation. In doing so, we have not aimed to reject Owen’s pluralist account of refugeehood, nor have we developed our own alternative position. Instead, we hope to have placed some pressure on the view in order to further develop the state system legitimacy view of refugeehood. It’s possible that the tripartite distinction between asylum, sanctuary, and refuge could be recovered. Doing so, however, would require deeper agreement on the concept of political membership itself.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ In-text page references correspond to Owen (2020).

² In earlier work, Buxton has critiqued Owen’s account and pushed for a shift in our understanding of membership. Here we hope to develop her critique by showing that Owen’s account of refugeehood is unstable precisely because of its ambiguity about membership, and also to lay bare how this interacts with the methodological commitments of legitimacy-based views.

³ Owen (11 fn14) refers to Beitz (2009) and Sangiovanni (2016) as influences.

⁴ Owen (58) writes that the ‘inoperative citizenship [of the sanctuary refugee] is not formally identical to the de facto statelessness of the asylum refugee, although it may still amount to a condition equivalent to de facto statelessness.’

⁵ One might adopt a broader conception of persecution, such the one developed by James Hathaway and Michelle Foster, who take persecution to be the ‘sustained or systemic violation of human rights demonstrative of a failure of state protection’ (Hathaway and Foster 2014, 185). The problem with this is that it inflates the notion of persecution beyond the distinctive role that it is supposed to play within Owen’s theory. If even these cases could count as persecution, then it is not clear that there would be a meaningful category of ‘sanctuary’ which could not be assimilated within that of ‘asylum.’

⁶ For a similar distinction, see Draper (2021).