Refugees, legitimacy and development

ABSTRACT In this paper I focus on Gillian Brock’s treatment of the case of refugees. After noting a potential distinction between our otherwise closely related theoretical approaches in which we view the refugee regime as a legitimacy repair (Owen) or legitimacy correction (Brock) mechanism, I draw a contrast between our ways of addressing this regime and argue that the difference between my historical approach and Brock’s presentist approach turns out to have implications for how we conceive what is due to refugees. Focusing on her advocacy of a developmental turn in refugee protection, I develop the concern that her articulation of this approach remains too closely tied to the humanitarian perspective of Betts and Collier in a way that underestimates the significance of political rights to refugee autonomy.

As Gillian Brock acknowledges, the theoretical arguments concerning the conditions of political legitimacy of the international state system that Joseph Carens and I had independently developed in addressing the international refugee regime are closely related to the arguments that she proposes and generalizes across the field of migration in relation to a defensible right of self-determination of states. Brock's picture of three sets of legitimacy conditions required for states to enjoy a right to self-determination makes explicit and develops what I take to have already been present in my own account in pointing to the international refugee regime as a legitimacy repair mechanism and making clear the duties of states to secure such legitimacy repair as a condition of sustaining the legitimacy of the international order of states on which their own claim to legitimacy is conditional (Owen 2016(Owen , 2020. Given the theoretical congruence of our approaches, it might be thought that I would have little to contribute to a critical evaluation of Brock's book. This is a presumption that I will, I hope, at least partly dispel through my focus on Brock's chapter on refugees. This discussion will have the useful function of both illustrating how our theoretical congruence need not entail convergence on commitments and in doing so highlights the role of contextualization with respect to the kind of practice-dependent theorizing in which Brock and I both engage.
My argument will proceed as follows. In the first section, I lay out Brock's account of legitimacy and its salience for reflecting on obligations to refugees. I turn, in section two, to Brock's approach to the case of refugees. In this section, I will suggest that Brock's analysis suffers from a lack of historical perspective with respect to the practice of refugee protection which I illustrate by contrasting her argument to my own similarly grounded approach which is grounded on a historical reconstruction of the development of the refugee regime and draws out distinctions between type of refugee protection -asylum for those fleeing persecution, sanctuary for those fleeing generalized violence, refuge for those fleeing a humanitarian crisis such as famine -on the basis of that historical reconstruction. In section three, I develop the implications of this contrast in arguing that while there are compelling reasons to see development-oriented approaches as part of a response to the global refugee situation, her analysis is not sufficiently attentive to the dilemmas that attend either refugee protection or this type of response to it. I conclude by considering the importance of integrating historical and explanatory analysis into normative reflection on refugee protection.

I
The normative framework that Brock proposes for her account focuses on articulating the legitimacy requirements of the right to self-determination of states within an international state system. The basic claim is that states' enjoyment of a right to selfdetermination is conditional on three requirements which we can gloss thus (Brock 2020, 38): LC1 Internal Requirement: respecting, protecting and fulfilling their own citizens' human rights. LC2 System Requirement: being part of a legitimate state system. LC3 Contribution Requirement: states' fulfilment of the positive obligations required for the cooperative project of sustaining a justified state system.
The key implications of this framework are: (1) When a particular state fails to meet the basic requirements of LC1 and LC3, this not only undermines the claim of that state to legitimately exercise rights of selfdetermination, it also undermines the legitimacy of the state system. (2) In order to address such legitimacy problems of the state system, this system needs to incorporate what Brock calls 'legitimacy correction mechanisms' (or what I had previously called 'legitimacy repair mechanisms'). Brock proposes that we may need different mechanisms for LC1 breaches and LC3 breaches. (Brock 2020, 39) One such legitimacy correction mechanism is the international refugee regime which secures the conditions of autonomy of refugees via the protection of states which are not their own and thereby repairs the legitimation gap created by their own states failure to secure their human rights. I have one initial issue to raise concerning Brock's elaboration of her framework before turning to the main focus on this article, namely, her approach to refugees. The issue that I want to draw attention to is that of system legitimacy. We can bring this into focus by reflecting on the following remarks: Individual state's performance varies greatly on whether they are meeting both internal and contribution requirements . . . While individual states may pass the threshold on (LC1) and (LC3), without significantly improved performance on (LC2) . . ., all states suffer from legitimacy defects. So, we have two important "weakest links" problems related to this issue. One problem is that no state can be fully legitimate unless we have a legitimate state system. Second, given that no state can be legitimate unless we have a legitimate state system, and the probability of at least one state not reach the legitimacy requirement is high, some might wonder why they should even aim to do what (LC1) and (LC3) require, because we are collectively going to fail, given at least one state is bound to let us down. (Brock 2020, 59-60) The implication of this passage is that, on Brock's view, system legitimacy requires that all states meet the legitimacy threshold. I have two objections to this view that may mark a difference between my own legitimacy-based account and Brock's.
The first is methodological. It is unreasonable to tie to system legitimacy to full compliance by states. Consider the remark by Joe Carens on refugees that Brock cites in discussing Carens and myself as 'fellow-travellers': The modern state system organizes the world so that all of the inhabited land is divided up among (putatively) sovereign states who possess exclusive authority over what goes on within the territories they govern, including the right to control and limit entry to their territories. Almost all human beings are assigned to one, and normally only one, of these states at birth. . . . Even if being assigned to a particular sovereign state works well for most people, it clearly does not work well for refugees. Their state has failed them, either deliberately or through its incapacity. Because the state system assigns people to states, states collectively have a responsibility to help those for whom this assignment is disastrous. The duty to admit refugees can thus be seen as an obligation that emerges from the responsibility to make some provision to correct for the foreseeable failures of a social institution. Every social institution will generate problems of one sort or another, but one of the responsibilities we have in constructing an institution is to anticipate the ways in which it might fail and to build in solutions for those failures. If people flee from the state of their birth (or citizenship) because it fails to provide them with a place where they can live safely, then other states have a duty to provide a safe haven. Thus, we can see that states have a duty to admit refugees that derives from their own claim to exercise power legitimately in a world divided into states. (Carens 2013: 196 cited in Brock 2020 Here Carens is, like Brock and myself, developing an argument that ties the legitimate exercise of state power to the legitimacy of the international state system, but for current purposes the key statement is this: 'Every social institution will generate problems of one sort or another, but one of the responsibilities we have in constructing an institution is to anticipate the ways in which it might fail and to build in solutions for those failures.' Carens' point is that social institutions, like human beings, are fallible and we need to acknowledge that fallibility in the construction of them through also constructing safeguards against their failure and mechanisms for dealing with failure what is happens. Legitimacy at the systemic level requires that we engage in such anticipation and building in of solutions, not that the social institution meets an unreasonable requirement of perfection such as every state being fully compliant with legitimacy requirements. The second is substantive. The point of 'legitimacy correction mechanisms' insofar as they function effectively is that they repair the immediate legitimacy gaps raised by states' that don't meet (LC1) and (LC3) criteria. Consider a world in which a particular state persecutes a minority group. This raise two problems: (a) the behaviour of the non-compliant state as a source of legitimacy problems and (b) the lack of human rights protection of the persecuted minority as the immediate legitimacy problem. Addressing (a) requires that legitimacy compliant states cooperate to address the conduct of the non-compliant state through governmental mechanisms (compatible with the grundnorms of the state system) which are aimed at bringing this state into compliance. Addressing (b) requires a legitimacy correction mechanism such as the international refuge regime which serves to repair the immediate legitimation gap opened the particular state's breach of (LC1) by ensuring protection for the persecuted minority. Insofar as the state system has appropriate governmental mechanisms for encouraging compliant conduct on non-compliant states and effective legitimation correction mechanisms for addressing the position of refugees, then the state system retains its legitimacy even where the rogue state remains stubbornly illegitimate. This is not a matter of 'interim' and 'full' legitimacy; rather the point is that although states cannot be fully legitimate unless the state system is legitimate, the state system can be fully legitimate even if not all states are.
With this preliminary difference between our common legitimacy-based approaches dealt with, let me turn to Brock's discussion of refugees.

II
Following Carens and myself, Brock sees the international refugee regime as a legitimacy correction mechanism aimed at ensuring protection of the human rights of refugees. As she writes: In many cases, especially of large populations of refugees such as, in recent years, the Rohingya Muslims or Syrian refugees, their basic human rights are seriously under threat. The state system has failed them. It therefore falls on the international community to act. But what should international agents do to assist? How should they go about trying to make arrangements for improved human rights protection under these circumstances? (2020: 113) Given Brock's sympathy for the practice-dependent argument concerning human rights that Beitz presents and on which she draws in constructing her theoretical framework, we might expect that she would at this point consider the purpose of the practice of refugee protection by considering how it developed and what was intended to do, and how we should evaluate its current form and failings against that backdrop. However, Brock does not engage in this reconstructive task. Rather she focuses her attention on one important type of case -mass refugee flows arising as flight from violent conflict such as that in Syria -and on one major proposal for restructuring the refugee regime to address this type of case, namely, the development approach advanced by Betts and Collier in their recent work Refuge: Transforming a Broken Refugee System (2017) which she largely endorses before briefly touching on a range of organizational and policy issues. This is, we may think, a curious way to proceed. She remarks that 'in order to know how to correct we must first analyse how our current arrangements are inadequate' (2020: 135) but surely in order to know how our arrangements are inadequate, we need to know what our arrangements should aim to do and for whom. Brock has a general and abstract response to this question: to protect the basic human rights of refugees, but in contrast to her treatment of the emergent practice of human rights, she does not make any effort to offer a reconstruction of the norms of the international refugee regime from consideration of its practice. Rather what she offers is a dive straight into what she sees as the failure of this regime. The question though is how we should understand this failure -and here there is some ambivalence in Brock's account.
We can start by going back to Brock's opening question: 'How should they [international agents] go about trying to make arrangements for improved human rights protection under these circumstances?' Notice that there is considerable ambiguity around the sense of 'these circumstances' in this question. In the paragraph in which this question is posed, Brock is pointing simply to the existence of 'a large gap between the grounds for endorsing the state system and the reality.' (2020: 113) If 'these circumstances' simply refer to these general conditions that require a refugee regime as legitimacy correction mechanism, then we might take Brock's focus on developmentbased approaches to refugee protection to represent what she sees as the appropriate response under conditions of general compliance by states with their obligations to refugees. This interpretation would be supported by her view that much of the existing literature has focused 'on individual state's obligations to offer refugees asylum at their borders and, especially, obligations to admit refugees for settlement or resettlement' and that 'such approaches ignore the possibility that thoughtful joint action might present more effective and sustainable ways of discharging our obligations well' (2020: 115). Yet, immediately before the paragraph in which she poses her question, Brock points to widespread non-compliance: 'Disturbingly, there is much evidence that states are increasingly denying asylum seekers the right to seek asylum by making entry to their territory impossible.' (2020: 112) Furthermore, immediately after posing her question, Brock highlights what she takes to be limitations in this context of noncompliance of the traditional 'durable solutions' of voluntary repatriation, local settlement in host states, and resettlement elsewhere. Read against this backdrop, we might take Brock to be offering development-based approaches as the best option available under circumstances of widespread non-compliance with the existing norms of the international refugee regime. The first issue, then, concerns the status of Brock's advocacy of development-based approaches. She claims: 'One final noteworthy positive feature of this approach I raise here is that it puts front and centre what constitute effective and feasible solutions to the deep problems surrounding refugee crises.' (2020: 137) I assume that these deep problems include the problem of non-compliance and hence Brock is proposing development-based responses as a way of effectively protecting refugee's basic rights that also encourages compliance by states. This does not mean that she may not also see such responses as the best response even under conditions, unlike our own, in which state compliance with duties of refugee protection is generally assured; indeed, I think this is the most plausible explanation for her position, namely, that she regards development-based approaches as the best option available even if states were generally compliant with existing norms.
If this is right, the 'failure' of the current regime is then to be seen as twofold. On the one hand, the traditional 'durable solutions' are inadequate on normative grounds even under conditions of general state compliance and we have independent normative reasons to expand the option set by focusing on development-based solutions because such solutions can better address the complex conjunction of refugee protection, host state development and post-conflict reconstruction. On the other hand, the existing regime is inadequate because it fails to motivate compliance among states, whether that be in the form of remote control policies designed to block access to territory in the Global North or in the form of denial of access to work, welfare or membership in the Global South, and development-based approaches can offer a way of negotiating this dilemma. This position would, I think, best make sense of her argument in which she seems to want to propose the developmental turn both on its own merits and as a response to contemporary conditions of non-compliance.
I will come back to the case of the developmental turn in refugee protection in the next section but before doing so I want to offer some reasons for thinking that Brock's strategy of argument has costs that matter for how we consider the international refugee regime and, given the proximity of our normative views, I'll develop this by reference to my own recent work on refugees which does attempt to offer a practice-dependent reconstruction of the international refugee regime.
This reconstruction begins by noticing the tension in contemporary refugee protection between humanitarian and political pictures of the refugee, showing how both of these pictures are rooted in the development of the practice of refugee protection, and proposing a normative reconstruction of the institution of refugeehood that does justice to what is valuable in each while overcoming them within a more comprehensive view which makes the norm of non-refoulement central to this regime and argues that, within the general category of refugees to whom the duty of non-refoulement is owed, we should distinguish between different types of refugee. Refugees, it is argued, are people for whom the international community must stand in loco civitatis, that is, as substitutes for their own state. What this relationship demands, it is proposed, can vary according to whether refugees require asylum, sanctuary, or refuge: Asylum is exemplified by the case of the people who have reasonable grounds to fear persecution by their home state (or by non-states actors from which their state is unwilling to offer protection). Today the Rohingya serve as one clear example of such persecution. Here we can endorse the view that a grant of asylum can be seen as expressing condemnation of the persecuting state (or the state not disposed to protect from persecution) 1 and the claim that central to asylum is the granting of a claim to membership. In 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 and when it is denied through persecution 2 to provide protection in a way that re-affirms the right to such standing -and this is what the legal status of asylum as a distinctive type of refugee status should be conceived as providing.
Sanctuary is exemplified by the case of people fleeing generalized violence and the breakdown of civil order. Many of those fleeing the civil war in Syria may fall into this category as would many refugees encamped in African states such as Kenya and Uganda. The claim advanced against the states to which such persons flee as representatives of the international community is a claim to sanctuary conceived as a space where one is protected against the threats to one's basic security, liberty and welfare posed by generalized violence and the breakdown of public order in one's home state without fear of being returned to that state insofar that the relevant conditions persist. In this context, the primary responsibility of the state that adjudicates the status of sanctuary-seekers is to ensure that they are subject to the norm of non-refoulement and have access to the basic security, liberty and welfare that the protective and enabling functions of citizenship would normally provide.
Refuge is exemplified by the case of people fleeing specific state failures such famine or natural disasters such as floods (where the line between 'state failure' and 'natural disaster' is typically blurred). The distinctiveness of the case of refuge is that it applies in the context of discrete and specific events where a person is so situated that they can secure themselves from the threat to their basic rights posed by the event in question by seeking immediate shelter across an international border and that this is their best reasonable option in the circumstances in which they find themselves. Grants of refuge thus act as to acknowledge and express a commitment to the basic rights of persons in the face of circumstances beyond the immediate control of their home state, and repatriation as soon as reasonable is the appropriate response. Refuge here serves the same basic function as international emergency assistance to persons displaced by the relevant events within the state and is essentially part of the same emergency assistance policy tool-kit. (Owen 2019 I argue that the legitimacy repair function of the international refugee regime requires such differentiation because distinguishing these three types of claim to refugee status as the distinct legal statuses of asylum, sanctuary and refuge is not only appropriately responsive to the distinct grounds of flight of the relevant persons (and hence discloses the distinct relations to persons that standing in loco civitatis can require) but is also essential to the proper functioning of the institution of refugeehood within global political society in its communicative role. Each type of status signals a distinct requirement to the international community not only in terms of action towards refugees but also action towards the state from which refugees flee (or to which persons are unwilling to return) which we can gloss as asylum: sanctions, sanctuary: conflict resolution, refuge: humanitarian assistance.
This historically grounded reconstruction of the normative character of the international refugee regime shares the same normative grounding as Brock's own argument and both positions broadly share the same view about the scope of refugeehood and its relationship to human rights, but there is a very clear contrast between Brock's general view and my own, namely, that she does not differentiate between responsibilities to refugees in terms of the grounds of their flight and the specific wrongs or harms to which they are subject. This, I think, flattens salient normative differences that a concern with legitimacy correction should acknowledge and address. It is a particularly notable difference that Brock's reflections do not address the importance for those subject to persecution of rapid acquisition of a new citizenship. However, since her primary focus is on the group that I identify as sanctuary refugees, drawing out our differences in this case may be most useful.
My own argument is that the loss of effective citizenship experienced by those who have fled conditions of generalized violence or the breakdown of public order can be conceived as the loss of reasonable conditions of effective social agency for which operative citizenship plays protective and enabling functions. To this experience of social powerlessness in the home state is then added the experience of social disorientation that comes with arrival in the unfamiliar social environment of another state. To repair this situation means providing refugees with conditions under which they can reasonably experience themselves as effective social agents, as agents who can make choices and plans about their futures that are not simply driven by the urgent requirements of practical necessity and who have some ability to shape the social environment in which those choices and plans are made. The basic requirements of this reparative responsibility could be expressed as provision of access to housing, health and welfare systems to protect sanctuary-refugees from the overwhelming demands of practical necessity; of access to opportunities for education, training or employment to enable sanctuary-refugees to make effective choices and plans about their lives; and of access to municipal political membership to enable sanctuary-refugees to experience themselves as having some say over the environment in which they are situated. Each of these provide routes of social integration for refugees within the state of sanctuary but the second and third are particularly important as involving the active participation of refugees in the social and political community of the locality in which they are situated. The ability to provide these kinds of basic requirements of social integration will thus also be an important consideration for the sharing of responsibility for sanctuary among states (where such sharing may take the form of some states acting as sites of sanctuary and others acting as supporters of sanctuary by providing skills and resources, for example).
Brock's view similarly stresses the importance of employment and education (2020: 119) and although she does not explicitly mention access to housing, health and welfare systems, I think that this is implicit in her view of the human rights protection that needs to be secured for refugees. The first point on which we may differ is, then, that of municipal political rights for refugees which I take to be integral to restoring refugee's sense of agency and for integration in the local community. Political rights do not appear at all in Brock's discussion except implicitly in her discussion of a pathway out of social limbo and this brings us to a second point of disagreement. Brock remarks: We cannot reasonably expect anyone to live in a state of high uncertainty about where they will live for an indefinite period of time. The basic idea is that after having been a refugee for some years, refugees' situations must be assessed so they can have more certainty about their future. . . . Those for whom it is judged that there is no credible prospect for repatriation in the foreseeable future should be offered a pathway to formal resettlement either in their haven country or another. (2020: 123) She suggests that after 5-10 years would be a reasonable time to have such an assessment and I will assume that 'foreseeable future' refers to up to 3 years ahead. In the case of a person who is deemed to have a credible prospect of repatriation, they can be 5-13 years with no political rights on Brock's view. In the case of someone who is not deemed to have a credible prospect of repatriation, they can be without political rights for 5-10 years plus whatever period of time it takes to qualify for citizenship in the state of settlement or resettlement. Is this reasonable? Refugees are vulnerable to arbitrary exercises of private and public power, and unlike voluntary migrants do not have meaningful access to the protections that may be offered by external citizenship rights (diplomatic protection and right to automatic re-entry to home state). This would seem to provide good reason for refugees to have access to naturalization on better (i.e. faster) terms that voluntary migrants (Owen 2013. It is, I think, a perplexing feature of Brock's account that it gives almost no attention to this issue at the levels of municipal or national citizenship. Indeed, if one adopts the kind of account of refugeehood proposed by Bender (2020) in which refugees are seen as those lacking public autonomy, then the salience of this issue will appear even more starkly. It is perplexing both because the issue of the political rights of refugees may straightforwardly be seen as part of 'restoring autonomy' (Brock 2020, 123) which Brock takes to be a central focus of refugee protection, and because, prima facie, political rights do not cut against the development focus that she favours. Let us then turn to her advocacy of developmental approaches to refugee protection.

III
The rationales for a turn to development-based approaches are twofold. The first is that the promise they hold out of combining the realization of several independently desirable features: the right to work and hence greater dignity and self-sufficiency for refugees, a lowering of the costs of (unproductive) humanitarian assistance, support for the development of the hosting state, and building capacities and skills in the refugee community that will support post-conflict reconstruction. The second rationale is that, under current conditions of non-compliance, the developmental turn offers a way of meeting the interests of Global North states in reducing the flow of refugees in their direction and the interests of the Global South states in economic development and achieving Sustainable Development Goals, and one that would represent a better use of funds than the currently highly skewed expenditure in which states spend far more on the 15% of refugees in the North than the 85% in the South. It thereby promises to encourage compliance and offer better value for money.
The case for this kind of approach has been most developed by Betts and Collier (2017) and their discussion heavily underpins Brock's argument. This is notable because Betts & Collier's argument is normatively grounded in a humanitarian duty of rescue combined with a commitment to autonomy (and thereby also a pathway out of limbo) and not on the kind of state legitimation view advanced by Brock. One effect of this is that their analysis focuses on promoting refugee autonomy in purely socio-economic terms, the issue of political rights only raises its head when the option of repatriation is off the table as part of their discussion of (re)settlement. That Brock's argument essentially reproduces the stance taken Betts and Collier (including their socioeconomic understanding of autonomy) raises the question of whether the state legitimation view makes any difference in this context. My own view, as I have noted in the preceding section, is that it does and that political rights are an integral element of the autonomy due to refugees. This concern is exacerbated if we note that Betts and Collier argue that onward migration from states of first safety should be managed and that if asylum is available to refugees in the state of first safety, then they have no refugeerelated grounds for further movement (although like anyone else they should be entitled to apply for visas -labour, family reunion, etc. -to move elsewhere). On this view, refugees as 'temporary residents' can be denied both voice and exit, the two classic mechanisms for expressing the claim that their interests are not being adequately addressed. Given how keen Betts and Collier are on the uses of game theory, it seems remarkably sanguine to assume that both that one can conceive of persons as autonomous in the absence of exit and voice mechanisms and that we can be confident of such 'autonomy' being secured by other actors. Perhaps the idea is that monitoring standards in the state by an international body will secure refugees interests by tying funding to standards, but this is at best paternalist. These are features of their approach which should, I think, concern Brock given the distinct normative basis on which her account rests in which public autonomy and not merely private autonomy is a central concern.
This also matters for a further reason. Betts and Collier's focus on development as an incubator of post-conflict reconstruction has a distinct tendency to treat refugees as materials for such reconstruction. Brock has a less instrumentalist attitude since she proposes that options to remain in the state of refuge should be offered 'to those who simply do not wish to go back' (2020: 131). However, she also expresses the concern relative to Syrian refugees who have settled in Germany that 'resettlement on Germany may work well for refugees and Germans who benefit from new citizens, but there can be an enormous net loss for those left behind in Syria that makes Syrians in Syria much worse off and considerably hampers their future well-being.' (2020: 136) This is an entirely reasonable concern although it rather ignores the roles that Syrians in Germany may play in contributing to postwar reconstruction (and in general Betts and Collier as well as Brock seem too easily to equate the ability to contribute to reconstruction with physical presence in a way that elides the benefits that a diaspora and carefully designed diaspora engagement policies may contribute). Notice though that it raises the question of what obligations citizens who have fled a state that has become a site of generalized violence following 'state collapse, dysfunction or civil war' (2020: 111 fn.1) have to that state or to their co-nationals that Brock does not address. Instead she relies on the empirical claim that the vast majority of refugees 'stay -and want to stay -in the region' (2020: 123) but even if true, it is not obvious that this expressed preference is not simply a function of preference adaptation under current circumstances in which the other choice is an expensive and dangerous trip to Europe.
None of this is to say that a focus on development is not welcome nor that the kinds of creative involvement of civil society actors are not worth pursuing nor, again, that the kinds of experimental results provided by Uganda's refugee work polices or Jordan's Special Economic Zone strategy cannot play important roles in the process of building well-designed polices in which refugee protection and agency is mobilized through development strategies. Like Brock, I would favour such creative experiments as one part of response to mass refugee flows but with two caveats. The first is that development-based approaches need to take seriously the political dimension of the autonomy of refugees and not simply limit their focus to socio-economic dimensions of autonomy. The second is to suggest that such an approach should only be one of a range of strategies alongside, for example, 'refugee matching' schemes and the use of existing diasporas. Moreover, while there is no doubt that, under appropriate conditions, voluntary repatriation can play a role alongside local integration in offering a durable solution for refugees, there is no need to regard to them as exclusive of each other within a development focus. Indeed, refugees who have acquired membership of the host state and thereby become dual nationals may reasonably feel more secure in exercising the choice to return to their state of origin given that they retain, as naturalized citizens, the right to re-enter the state of sanctuary at any time, while their connections to the state of sanctuary also provide a resource for building stronger links between the two states (whether in terms of economic, cultural or social relationships) to the benefit of each. Furthermore, in the likely event that some sanctuaryrefugees choose to return while other choose to stay, the resulting transnational relationship of home state citizens and sanctuary state diaspora can, given supportive policies in these states, similarly serve as a resource that works to the developmental advantage of both states and further support, and be supported through, regional forms of governance. It is worth noting here two wider trends in contemporary global politics, namely, the growth of regional bodies that enable greater mobility and interaction between states and the 'transnationalisation' of states in terms of the spread of dual nationality as a permitted status, the normalization of expatriate voting rights, and the intensification of diaspora engagement policies. These trends suggest that integrating political rights into development strategies and allow refugees a wider range of options than being held immobile in neighbouring states could strengthen the development turn in a way that is more consonant with the normative basis of the state legitimacy approach that Brock and I share.

Conclusion
I have argued that Brock's focus on development represents a valuable route for thinking about refugee protection but that there is a tension between the views of Betts and Collier grounded on a humanitarian duty of rescue and the normative stance that Brock and I advance in which refugee protection is seen as a legitimacy repair mechanism. I have suggested that Brock's analysis for all its welcome proposals is not sufficiently attentive to this tension and in particular to the importance of political rights as a key part of refugee autonomy -and that this is a product of the fact that she does not extend her practicedependent approach to a reconstruction of the point and purpose of the international refugee regime. Brock's work, in this book and elsewhere, has a very welcome tendency to engage with the salient empirical literature in order to address the real problems and challenges facing the global community, however, the immediacy of her focus can also, I am suggesting, have costs and I have tried to illustrate that point here.

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