Justice and Internal Displacement

This article develops a normative theory of the status of ‘internally displaced persons’. Political theorists working on forced migration have paid little attention to internally displaced persons, but internally displaced persons bear a distinctive normative status that implies a set of rights that its bearer can claim and correlate duties that others owe. This article develops a practice-based account of justice in internal displacement, which aims to answer the questions of who counts as an internally displaced person and what is owed to internally displaced persons (and by whom). The first section addresses the question of who counts as an internally displaced person by offering an interpretation of the conditions of non-alienage and involuntariness. The second section articulates an account of what is owed to internally displaced persons that draws on and refines the idea of ‘occupancy rights’. The third section sets out an account of the role of the international community in supplementing the protection of internally displaced persons by their own states.

(IDPs) emerged (Cohen and Deng, 1998a). International legal principles concerning the treatment of IDPs have been articulated and codified in international legal frameworks such as the Guiding Principles on Internal Displacement (Deng, 1999), the Great Lakes Protocol (2006), the Kampala Convention (2009) and the Inter-Agency Standing Group (IASC) Framework on Durable Solutions for Internally Displaced Persons (IASC, 2010).
Political theorists, however, have paid little attention to internal displacement. Most work on forced migration in political theory has tended to focus on movement between states and has tended to ignore movement within states. As Alex Sager (2018: 20) puts it, political theory often 'treats the mobility of people within the boundaries of the state as irrelevant'. While there has been a significant focus in political theory on refugee status, there has been no comparable systematic examination of the IDP status -despite the fact that, at end of 2019, there were 45.7 million IDPs, as compared with 26 million refugees (UNHCR, 2020: 2). This narrow focus on refugee movement has obscured the normative significance of internal displacement. IDPs bear a distinctive normative status that implies a set of rights that its bearers can claim and correlate duties that others owe. To critically evaluate the ways in which IDPs are treated in practice, we need a normative theory that tells us who is eligible for IDP status, what rights those people have in virtue of that status, and what correlate duties other agents owe.
This article takes up the task of developing such a theory. I take a practice-based approach, which seeks to normatively reconstruct the concept of the IDP (see Sangiovanni, 2008Sangiovanni, , 2016. I aim to answer the questions of who counts as an IDP and what is owed to IDPs (and by whom). The first section addresses the question of who counts as an IDP by offering an interpretation of the conditions of non-alienage and involuntariness. The second section articulates an account of what is owed to IDPs that draws on the idea of 'occupancy rights'. The third section sets out an account of the role of the international community in supplementing the protection of IDPs by their own states. The account that I develop begins with the practices of IDP protection, but it does not end with them: its aim is to articulate a normative standard for critique and reform of those practices.

Who Counts as an IDP?
According to the Guiding Principles on Internal Displacement: Internally displaced persons are persons or groups of persons who have been forced or obliged to flee or to leave their homes or places of habitual residence, in particular as a result of or in order to avoid the effects of armed conflict, situations of generalized violence, violations of human rights or natural or human-made disasters, and who have not crossed an internationally recognized State border. (Deng, 1999: 484) The two 'distinctive features' of the IDP identified here are that 'movement is involuntary or coerced and that the populations affected remain within their national borders' (Cohen and Deng, 1998a: 16). Each of these features -involuntariness and non-alienage -plays an important role in specifying who counts as an IDP.

(Non-)Alienage
We can begin with the concept of (non-)alienage: the fact of whether or not an individual is outside of their state. Those who are inside their state are in a condition of non-alienage, whereas those who are outside the territory of their state are in a condition of alienage.
At first blush, non-alienage seems to be an obvious and central aspect of internal displacement. It is worth interrogating its moral relevance for the ascription of IDP status, though, because our view on the relevance of non-alienage for IDP status will bear on our view on a conceptually related question about the role of alienage in refugee status. In legal practice, alienage is a necessary condition for refugee status. 2 Some, however, have argued that alienage is not necessary for the justified ascription of refugee status from a normative point of view (e.g. Beaton, 2020;Owen, 2016: 279-80;Shacknove, 1985). If they are right, then some who are displaced within the territory of their state may well have a claim to refugee status, and we will need to know how we can distinguish such 'internal refugees' from IDPs. Or, if we are to show that they are mistaken, then we will need to explain the moral significance of non-alienage.
In legal practice, refugees must also have a 'well-founded fear' of persecution on the basis of certain protected characteristics, and so we might think that we can distinguish IDPs through the absence of persecution. 3 The absence of persecution, however, will not suffice to distinguish IDPs, for two reasons. First, the normative relevance of persecution condition is itself heavily contested, so to hang the distinction between IDPs and refugees on the persecution condition would be to adopt a controversial position in this debate. 4 Second, relying on the persecution condition would make us unable to distinguish IDPs from those eligible for 'complementary protection'. Complementary protection is an auxiliary form of international protection enjoyed by those who cannot appeal to their own state for protection but who do not meet the persecution condition (Lister, 2019;McAdam, 2007). For example, many of those who fled from Somalia after the collapse of Siad Barre's authoritarian regime in 1991 were not fleeing direct persecution, but rather conditions of state breakdown and generalised insecurity (Betts, 2013: 135-136). A narrow view of refugeehood that takes persecution as a necessary condition would not identify such people as refugees, but we presumably still want to be able to distinguish these people from IDPs. To do so, we cannot rely on the persecution condition, and so we will need to assess the moral significance of non-alienage itself.
My contention is that non-alienage is not itself the morally relevant feature of the IDP. But this does not mean that there is no morally significant distinction to be drawn between IDPs and refugees (and other forced migrants with claims to international protection). 5 On what I will call the relational view, the morally relevant feature that distinguishes the IDP from the refugee is the status of the relationship in which she stands with her state. Non-alienage, on that view, is best interpreted as an imperfect proxy for tracking the status of that relationship. Refugees cannot claim the protection of their own state because their relationship with their state has comprehensively broken down. IDPs, however, can claim protection from their own state. Their relationship with their state remains fundamentally intact, even if it is to some degree compromised. The relational view thus leaves open space for the possibility of 'internal refugees', but preserves an important distinction between IDPs and refugees.
When one's location matters, it ordinarily matters because it marks out some other morally relevant feature, such as the relations of jurisdiction in which one stands. One important feature of the contemporary international order is that jurisdictional authority is largely organised on a territorial basis. Another is that states are charged with the primary duty of respecting, promoting and fulfilling the human rights of their own members, 6 and have standing to act in that capacity (Karp, 2020). 7 Ordinarily, these two features work together, in that states discharge their duties to respect, promote and fulfil the human rights of their members by exercising jurisdiction over the territory in which their members are located.
States' territorial jurisdictions and their human rights-protecting relationships with their members are not co-extensive, however. States' duties persist even when their members are no longer under their territorial jurisdiction. When a tourist is on holiday in another state, her home state still has a duty to ensure that her human rights remain protected over time -for example, by 'giv [ing] her documentation for purposes of international travel, [. . .] stand[ing] up for her in disputes with nations in which she is travelling or residing, [. . .] proving other forms of "diplomatic protection"' (Aleinikoff and Zamore, 2019: 35) -even if the state she is visiting has a duty not to violate her human rights.
At the same time, states can lose their standing to act as the primary guarantor of their members' human rights, even within their own territory. Where states continually fail to secure the human rights of some of their members, they can lose their standing as the 'first-level' guarantor of those members' human rights and the international community has a 'second-level' duty to step in to provide protection in their place (Beitz, 2009: 109). In the limit case, the international community may be permitted (or even required) to encroach upon the territorial jurisdiction of the state, for example, to prevent ethnic cleansing, mass expulsion or enslavement (see Cohen, 2014: 159-222). Of course, not just any human rights deficit will mean that the state loses its standing. Sometimes, the state will be basically competent and acting in good faith, but will be unable to secure the human rights of all of its members all of the time. In these cases, the international community's primary responsibility is to 'supplement' the protection of those members' human rights, rather than to 'replace' the state (Owen, 2016: 279-280).
We can distinguish, then, between cases where the human rights-protecting relationship between the member and the state has broken down, and cases where it remains fundamentally intact, even if it is compromised to some degree. A useful way of illustrating this difference is by analogy with a distinction used in a different context by Tamar Schapiro, who distinguishes between an 'offence' and a 'betrayal' against a moral relationship: An offense issues from the standpoint of one whose basic commitment to the relationship is not in question. As such it has a bearing on the degree of perfection of the relationship, but it does not undermine the relationship's basic integrity. A betrayal, by contrast, issues from the perspective of one who is legitimately subject to the demands of the relationship, but whose fundamental commitment to the relationship is in question. As such, betrayals throw the basic character of the relationship into question (Schapiro, 2006: 53). This analogy is not perfect, since the 'betrayal' language implies that the state must intentionally disregard its members' human rights rather than merely chronically fail to secure them. But it is helpful in mapping the conceptual terrain of the member-state relationship, in that it shows that there can be cases where particular human rights go unfulfilled without the moral relationship between the member and the state being fundamentally broken. The status of this relationship does not map perfectly onto the territorial jurisdiction of the state; it can remain intact when the individual is outside of the territory of the state, and it can break down when the individual is within the territory of the state.
To see why it is the status of this relationship that is of fundamental moral significance in drawing the distinction between IDPs and refugees, rather than the displaced person's territorial location, consider two cases in which these come apart.
First, consider a case in which the international community protects those who are within their country of origin. Since the early 1990s, the United Nations High Commissioner for Refugees' (UNHCR) 'Comprehensive Refugee Policy' has seen it become increasingly operationally involved in states that are producing refugees, often by setting up 'safe zones' within conflict countries and providing protection designed to pre-empt the need for flight (Suhrke and Newland, 2001). If we took (non-)alienage to be fundamental in drawing the distinction between IDPs and refugees, then those being protected by UNHCR in safe zones would not count as refugees, since they do not meet the alienage criterion. But such people cannot appeal to their own state for protection, and we can expect that they would flee across borders (and have a claim to refugee status) were it not for UNHCR's provision of protection in situ. They have a compelling claim to protection of the sort that only the international community can provide. If we take the member-state relationship to be fundamental in drawing the distinction, we can explain this judgement by pointing to the fact that these people's relationships with their state have broken down, even if they remain within its territory.
Second, consider a case where an individual is displaced across a border, but where her state retains its standing as the guarantor of her human rights. In April 2014, when the banks of the Mamore River burst after heavy flooding in Bolivia, some 120 families in the border region of Guayaramerín fled across the border and were sheltered in Brazil, as floods prevented them from being easily moved to Bolivian shelters (Sedeh, 2014: 179-180). Suppose that if Brazil failed to provide adequate protection, Bolivia would have robustly fulfilled its obligations of diplomatic protection and would have provided protection within Bolivia if necessary. In this case, it seems clear that these families, despite being displaced and being outside of their own state, still stand in the right kind of relationship with their state for us to describe their state as the guarantor of their human rights. It would be strange to treat such people as refugees or claimants of international protection, even though they are displaced outside of the territory of her state. After all, it was merely for reasons of efficiency that their protection took place in Brazil. The agent with the primary duty to ensure the protection of their human rights remained Bolivia, even if the responsibilities associated with this relationship had, in this case, been informally transferred to Brazil. If we take the status of the member-state relationship to be morally basic in drawing the distinction between IDPs and refugees, we can explain this judgement by appealing to the fact that the member-state relationship remained intact.
In each of these cases, our judgements about who is entitled to the international protection associated with refugee status come apart from whether or not those involved have left the territory of their state. The best explanation of this is that it is the status of the member-state relationship, not (non)-alienage itself, that is morally fundamental in drawing the distinction between IDPs and refugees. This latter case also opens up conceptual space for the possibility of an IDP who is outside of the territory of her state, which may appear counter-intuitive. It is worth recognising, however, that my focus here is on the underlying philosophical basis upon which the rights associated with IDP and refugee status should be granted, and that non-alienage may still play a role in the practice of IDP protection. As we can see from this case, it is rare for an individual to be displaced outside of the territory of her state and for her state to be in a position to protect her human rights. Generally, a state needs to have jurisdiction over the territory in which the individual is displaced to discharge its duties. This means that, as a rule, we have good reason to treat those outside of the territory of their state as being eligible for international protection.
(Non-)alienage, then, has some practical significance as a proxy for identifying IDPs and refugees. It is not, however, itself the morally relevant feature of either IDP or refugee status. Analogously, consider the legal age requirement for getting a tattoo. Presumably, it is one's maturity, or one's ability to conceptualise the long-term consequences of getting a tattoo, rather than the number of times the earth has rotated around the sun since one's birth, that is morally relevant for being ascribed the right to get a tattoo. Age, though, generally works tolerably well as a proxy, which is a lot easier to operationalise (and a lot less invasive) than a direct assessment of maturity.
The relational view has critical potential, though, because (non-)alienage is not always a good proxy, particularly when it comes to those within the state. In some cases, which have traditionally been viewed as cases of internal displacement, particularly in conflict situations, states do appear to have lost their standing to act as the guarantor of their members' human rights. Hathaway suggests that a significant source of the relative popularity of the IDP label among states is that it was a convenient way for states to shirk their responsibilities to would-be refugees in conflict situations. As he points out, '[t]hese persons would in most cases have qualified for refugee status had they not been encouraged, and at times compelled, to remain inside their own country' (Hathaway, 2007: 356). Often, the establishment of 'safe zones' within conflict countries has served this function (see Long, 2013). In cases like these, taking the member-state relationship to be morally fundamental has the advantage of enabling us to explain why these displaced people, even if they have remained inside their state, are entitled to the more robust forms of international protection associated with refugee status.
The relational view thus implies a revision of the practices of IDP protection. It means that when we ascribe IDP or refugee status, we make a judgement about whether or not the individual concerned is entitled to international protection or whether they should appeal to their state to protect their human rights. Civil conflict will often (but not always) be indicative that a state has lost its standing to act as the ongoing guarantor of its members' human rights, at least vis-à-vis some segment of its membership. Other forms of displacement (such as disaster displacement) will often be a matter of a state trying to uphold its responsibilities in good faith, but facing exigent circumstances which mean that its capacities are under pressure. Historically, however, many of the cases that have motivated the development of the IDP protection regime have been cases of civil conflict. 8 On the relational view, it may be more appropriate to treat those who are displaced in such cases as refugees, rather than IDPs.

Involuntariness
The second distinctive feature of the IDP is that IDPs have been 'forced or obliged' to move: their movement is involuntary. Unpacking the conception of involuntariness at work in the practices of IDP protection can also help us to clarify who counts as an IDP.
We can distinguish between two conceptions of involuntariness. One conception draws inspiration from Joseph Raz's conception of autonomy. Raz argues that to undertake genuinely autonomous action, individuals must have an 'adequate range of valuable options' available to them (Raz, 1988: 373-377). If an individual's option set is overly constrained, then her choice is not genuinely autonomous, and so it cannot be properly thought of as voluntary. 9 Valeria Ottonelli and Tiziana Torresi appeal to a similar idea in articulating a standard for voluntariness in migration: 'a migration project is voluntarily undertaken only if the available alternatives at home are good enough for the migrant' (Ottonelli and Torresi, 2013: 798). In the context of internal displacement, we might say that movement is involuntary, and that those on the move thus have a claim to IDP status, when movement is selected from an inadequate range of options. This provides us with a rough standard for determining when movement is involuntary. Call this the broad view.
On the broad view, a significant number who are on the move would count as IDPs. Consider, for example, those who migrate to cities to take up waged labour in response to environmental degradation or land use changes that threaten the agricultural practices upon which they rely. Or consider those who move in response to transformations in housing markets which make remaining in place unaffordable (see Huber and Wolkenstein, 2018). Even without specifying in detail what counts as an adequate range of valuable options, it seems clear that these people appear to have few valuable options available to them. According to the broad view, their movement is involuntary, and so they should count as IDPs.
A second conception of involuntariness can be called the narrow view. On the narrow view, involuntariness is understood in a more specific sense for specifying who should count as an IDP: in the sense of it being reactive. The idea of reactive movement comes from Anthony Richmond's (1993) sociological theory of migration, in which he distinguishes between proactive and reactive movement. Reactive movement is precipitated by sudden events and changes, which 'disrupt the normal functioning of the system [upon which those affected depend]' and 'destroys the capacity of a population to survive under the prevailing conditions' (Richmond, 1993: 16). Proactive movement, by contrast, is typically planned and anticipated, even though the options available to those on the move may be structured by a range of political, economic, environmental, social and bio-psychological 'structural constraints' (Richmond, 1993: 12, 15-17).
My claim is that for the purposes of an account of the normative status of the IDP, we should adopt the narrow view. Before substantiating this claim, however, it is worth noting that the narrow view does not claim that all proactive forms of movement are voluntary. It seems entirely sensible to suggest, for Razian reasons, that movement in response to slow-onset economic decline or environmental degradation is not voluntary. The narrow view just says that one specific kind of involuntary movement -reactive movement -is the proper concern of the IDP protection regime. Those affected by other forms of involuntary movement may well have important normative claims with respect to that movement. My claim is only that we should be concerned specifically with reactive movement for the purposes of an account of the normative status of the IDP.
There are two reasons to prefer the narrow view over the broad view. The first reason is that the broad view's capacious understanding of involuntariness fits poorly with the practices of IDP protection. The causes of internal displacement identified in the Guiding Principles -armed conflict, generalised violence, human rights violations and disasters -provide central examples of the kind of movement that the IDP regime is suited to address: the movement of those who find themselves suddenly uprooted from their homes. The inclusion of poverty-driven migration was considered in the drafting of the Guiding Principles, but was rejected on the basis that there were 'distinct protection and assistance needs resulting from forced [i.e. reactive] displacement' and that 'to expand the definition would risk losing this focus' (Mooney, 2005: 13). The idea that reactive movement is the specific concern of the IDP protection regime is also indicated in an earlier working definition of the IDP formulated in the drafting process, which identified IDPs as those who flee 'suddenly or unexpectedly' (Mooney, 2005: 10). This language was eventually dropped because it had the potential to exclude cases where states planned and perpetrated displacement, as in the case of the targeted uprooting of the Kurds in Iraq in the 1970s and 1980s (Cohen and Deng, 1998a: 17). The spirit of this clause, however, is that displacement which involves the rapid disruption of background conditions of stability is importantly distinct from proactive forms of involuntary movement in which individuals make migration decisions from among a constrained set of options. These different kinds of involuntary movement demand different kinds of responses. Proactive forms of involuntary movement may be better governed through anticipatory forms of assistance, such as development assistance, labour migration regulation, or planned relocation programmes. If we take seriously the particular purpose of the IDP protection regime, then we should restrict our understanding of the involuntariness of internal displacement to the more specific sense of reactive movement to stop us from inflating the concept of the IDP.
A second reason concerns the distinctive nature of the harm of reactive displacement. The IDP protection regime's specific focus on reactive displacement can be vindicated by the fact that the narrow view captures a distinctive harm that is not captured by the broad view. We can see the role of the IDP protection regime as being to respond to this distinctive harm.
The harm of reactive displacement can be understood in terms of the interest that we have in being able to form and pursue our own life-plans against stable background conditions. Anna Stilz ( , 2019 and Margaret Moore (2015) both point towards this interest in their accounts of what they term 'occupancy' and 'residency' rights, respectively. Both Stilz and Moore begin with the identification of what they take to be a central human interest: the interest in being able to form and pursue one's own life-plans (Moore, 2015: 38;Stilz, 2013: 335; see also Ypi, 2012: 295). We depend on background conditions of stability that are located in the shared and private spaces upon which we depend. Life-plans are located in the sense that they are organised around our 'expectations of continued use of, and secure access to' (Stilz, 2013: 335) the spaces in which we participate in social practices. Beyond these meso-level shared spaces, we also depend on secure access to micro-level spaces such as the home. Secure access to housing is a basic prerequisite for living an autonomous life, since without it individuals would be 'constantly having to negotiate changes in their place of habitation' which would 'undermine their ability to develop and exercise a plan of life' (Wells, 2019: 410; see also Nine, 2018). As Moore (2015: 38) puts it: 'our individual plans and pursuits depend on a stable background framework, and this is provided by security of place'.
Both reactive and proactive displacement set back our interest in a located background conditions of stability against which we form and pursue life-plans, but they do so in importantly different ways. Neither Stilz nor Moore distinguishes between reactive and proactive movement, but there is an important difference between the two in terms of their effects on the background conditions upon which we depend. In the case of reactive displacement, sudden disruptions destroy the stable background conditions against which we form and pursue our life-plans. These disruptions upset the expectations that we have to continue to be able to securely use and access particular spaces, including our homes, which enable us to form and carry out our plans. Beyond the discrete losses involved in any particular instance of displacement -of property, land, access to kinship networks, and so on -it is the stable background upon which we depend that we lose. To recover our ability to form and pursue life-plans, we need to re-establish a stable background framework that enables us to form and pursue life-plans. The ability to form and pursue one's own life-plans is also implicated in situations where there is a pervasive and significant risk of reactive displacement. Where the background conditions of stability upon which we depend may collapse or be swept out from under our feet at any moment, it becomes difficult to plan our lives. Jonathan Wolff and Avner De-Shalit call this phenomenon 'planning blight' (Wolff and De-Shalit, 2007: 69).
In the case of proactive movement, slow-onset changes also affect our ability to form and pursue life-plans, but they do so by gradually eroding the background conditions upon which we depend, rather than by eradicating the foundations upon which our plans rest. The gradual erosion of the background conditions forecloses options that may figure in our developed or anticipated plans. In this case, the changes to the background conditions alter and constrain the options available to us. But rather than needing to re-establish a stable background against which we can develop any plans at all, we can respond to these slow-onset changes either by revising our plans and adjusting our expectations, or by working to prevent or alter the changes in circumstances that threaten our existing plans. This does not mean that those affected by proactive displacement do not suffer a harm: the foreclosing of the options available to them is certainly harmful, and it may be no less significant than the harm of reactive displacement. The crucial point is rather that this harm is of a qualitatively different kind, and that it warrants a different response. 10 This provides us with a principled basis for adopting the narrow view for determining who counts as an IDP: the narrow view captures a distinctive harm, the destruction of the located background of stability upon which our life-plans rest. The IDP protection regime's function is best interpreted as being to address this specific harm.
Taken together, these interpretations of the concepts of non-alienage and involuntariness can provide us with an account of the normative status of the IDP. IDPs are those who are reactively displaced from their homes, but whose relationship with their state remains fundamentally intact, even if imperfect. This account of who counts as an IDP enables us to distinguish IDPs from both refugees, whose relationship with their state has broken down, and from other involuntary migrants, such as those moving in response to slow-onset environmental degradation or chronic poverty. It thus provides us with a critical standard for assessing who has a claim to protection as an IDP.

What Is Owed to IDPs?
The conception of the involuntariness developed in the previous section can also help us to clarify the content of the duties that we owe to IDPs. Since IDP status picks out those who face the harm of reactive displacement, the rights and correlate duties associated with IDP status can be unpacked by reference to justified claims that IDPs have to be protected against that harm. To say that these are 'rights' is to say that these correlate duties have pre-emptive force, which precludes us from considering the merits of protecting the interest at stake on a case-by-case basis (Raz, 1988: 186;Stilz, 2013: 341). To say, further, that these are 'human rights' is to say that they are 'requirements whose object is to protect urgent individual interests against certain predictable dangers ('standard threats')'; that states have a duty to respect, promote and fulfil such rights; and that, when they fail to do so, this is a matter of international concern that provides reasons for action on the part of the international community (Beitz, 2009: 109; see also Shue, 1996: 34). Displacement is a 'standard threat' to the urgent individual interest that we have in a located background of stability which, as I argue in this section, justifies a particular set of rights and correlate duties on the part of states. In the third section, I examine the international duties associated with IDP protection.
In their discussion of occupancy rights, Stilz and Moore argue that the right to occupancy has two components: first, a liberty to reside in a particular space, and second, a claim against removal from that space (Moore, 2015: 36;Stilz, 2013: 327-328). Moore and Stilz are primarily concerned with coercive expulsion, but we can employ the concept of occupancy rights to clarify the duties that states have with respect to internal displacement. To do so, though, we need to extend the concept of occupancy rights. Stilz contends that occupancy rights only extend so far as to justify a claim-right against being removed from a particular place. Reactive displacement, however, can be precipitated by purely natural disasters and need not involve any agent infringing upon the claim-right against being removed. My claim is that occupancy rights imply a correlate positive duty on the part of the state to protect and maintain the background conditions of stability upon which we depend, and to restore them when they are disrupted, rather than merely a negative duty not to disrupt background conditions of stability. This means that even if events which precipitate displacement arise naturally without any infringement of a claim-right against being removed, states may commit a wrong if such events lead to violations of the rights of their members with respect to displacement.
The precise extent of this duty should be made clear. It would not be reasonable to expect states to prevent all cases of displacement which arise under any circumstances. Displacement is hard to predict, occurs suddenly, and may be of such scale that it is not feasible to stop everyone from being displaced. Despite our best efforts, displacement may still, tragically, occur. In such cases, I do not think we can say that those affected have been wronged, even if they have been harmed. But we can still specify the rights that individuals have against their state have with respect to displacement. On my view, such rights include, first, a right to the reduction of the risks of displacement to tolerable levels, and second, a right to have the background conditions of stability restored quickly and effectively when displacement does occur. These rights entail correlate duties on the part of others (in the first instance, states) to reduce the risks of displacement and restore the background conditions of stability when it does occur.
The first right is not strictly something that is owed to IDPs, but is rather a right that each individual has against her state to make reasonable efforts to prevent her from becoming an IDP. Perhaps ideally states would reduce the risk of reactive displacement to zero, but this will not be possible or reasonable in most circumstances. A threshold of 'tolerability' is more sensitive to the fact that it is not practically possible to entirely eliminate all risk of reactive displacement, and reflects the idea that low levels of risk need not disrupt our ability to form and pursue life-plans. What counts as a 'tolerable' level of displacement risk is hard to specify in the abstract and may well vary across circumstances. It should, however, be low enough to enable to reasonable individuals to depend upon a background of stability. In practice, protecting this right will require states to do things such as invest in disaster risk reduction and climate change adaptation programmes.
The second right can only be fulfilled through the provision of assistance once displacement has occurred. We must recognise that our best efforts to protect against displacement may fail, and that where they do, we have a duty to restore the background conditions of stability that enable us to form and pursue our own life-plans. This duty is owed to IDPs qua IDPs: even if the state discharges this duty perfectly, those to whom it is owed still count as IDPs (at least until their reactive displacement has been fully addressed). There are broadly three ways in which this duty can be discharged: return, local integration and resettlement. In international legal practice, these are taken to be the 'durable solutions' to internal displacement, and have been codified most clearly in the IASC (2010) Framework.
Each of these options can restore the background of stability in its own way. Often, return will be the preferable solution. It may, for example, enable IDPs to reconstitute their original plans by reclaiming property and lands. It may also have additional significance in terms of preserving IDPs' ability to participate in valued cultural or religious practices with specific connections to particular places, or enabling them to assert their status as equal members of the community from which they have been displaced (Bradley, 2018: 235-240). The idea that those who have been expelled from their lands and homes have a right to return enjoys broad support among theorists of territorial rights, 11 and this can be fairly straightforwardly extended to IDPs if we accept that states have a positive duty to protect their members against displacement. In some cases, however, return may not be possible or desirable. For example, a natural disaster may mean that there is a lack of habitable land. Or, in more protracted situations of displacement, those affected may begin to develop new life-plans and expectations in the environments in which they find themselves. In such cases, local integration may be a preferable way of nurturing such plans and expectations, by ensuring that they rest on stable background conditions. Where local integration is not desirable, resettlement may be preferable. Resettlement projects for IDPs, if undertaken well, will allow IDPs to exercise some measure of choice over the new environment in which they will live. They may enable IDPs to move to a new environment which provides particular goods (e.g. resources or access to culturally valued practices) that may enable them to pursue their own life-plans.
The best way to restore background conditions of stability is likely to depend on particular circumstances of different IDPs, and in international legal practice IDPs are supposed to enjoy the 'right to choose freely between return, local integration, or resettlement' (Kälin, 2008: 125). One rationale for this is that IDPs know their own circumstances best, and so will often be best-placed to know which 'durable solution' best fits their circumstances. But this is not the only or best rationale. Even if IDPs are mistaken about which option best suits their circumstances, they should still have the power to decide between these options. This reflects the idea that we value not only the life-plans that we happen to have, but also our ability to autonomously form and pursue our own life-plans (see Stilz, 2019: 41-43).
Taken together, these two rights specify the content of the state's positive duty to protect its members against the standard threat of displacement. Stilz, however, has argued that although there is a negative duty not to remove others, occupancy rights do not always justify a positive duty to protect the background conditions of stability which enable us to form and pursue our life-plans. She writes that '[t]he fact that we ought not to interfere with others' territorial occupancy, then, does not necessarily entail that we are also obliged to subsidize them in maintaining their located life-plans' (Stilz, 2013: 344). Stilz uses the example of a mining community, and argues that its members do not have a right that others subsidise their economic activity to prevent their life-plans from being disrupted (Stilz, 2013: 343-345). Others, she argues, 'have strong countervailing interests in not bearing the burdens required to maintain the miners in their current occupations'. Stilz does accept, though, that others may have a duty to 'cushion dislocation' in the mining community through social welfare benefits or worker retraining (Stilz, 2013: 344).
If Stilz is right about the mining community, then we might think that a positive claim to protection against displacement is similarly unjustified, since it would involve imposing burdens on others to pay for the reduction of displacement risks and for the provision of displacement assistance when it does occur. More than requiring mere non-interference, protection against displacement requires positive protection of the conditions of stability that make forming and pursuing life-plans possible.
I agree with Stilz that we must take into consideration the burdens imposed on others in determining whether or not a right can be justified. 12 There are, however, important differences between the mining case and the case of protection against the standard threat of displacement. The most significant difference is that the countervailing reasons to which Stilz appeals in the mining case do not obtain in the displacement case. Stilz argues that the benefits which accrue to citizens as a result of having a competitive market economy ('dynamic innovation, lower consumer prices and greater opportunities') outweigh the miners' interests in being insulated against market competition (Stilz, 2013: 344). Protection against reactive displacement, however, does not require us to deny citizens the benefits of a competitive market economy by intervening to prevent economic competition. It requires only that some portion of the state's resources be used to address the threat of displacement through programmes of disaster risk reduction and reactive assistance where displacement does occur. Protection against displacement is more like the social welfare benefit provision that Stilz finds acceptable in the mining case than the intervention in the competitive market economy that she finds unacceptable. A positive duty on the part of the state to protect and restore the located background conditions of stability should as such be relatively uncontroversial; it is like the other kinds of social insurance mechanisms that states put in place to protect the basic interests of their members.

Who Owes What to IDPs?
Thus far, our focus has been on the rights that the displaced have against their own state. In practice, the state is the agent charged with the primary responsibility for the protection of IDPs, as is indicated in Principle 3(1) of the Guiding Principles: National authorities have the primary duty and responsibility to provide protection and humanitarian assistance to internally displaced persons within their jurisdiction. (Deng, 1999: 485) However, the governance of internal displacement as it is practised also has an international component. UNHCR has a large role in delivering IDP protection, but so do a wide range of other international agencies (Phuong, 2005: 92-102). Internal displacement is taken to be a matter of international concern, and when states cannot protect their own members against the standard threat of displacement on their own, they call on the international community to supplement their efforts. An adequate theory of the normative status of the IDP will need to explain what duties, if any, the international community has with respect to displacement.
According to what I call the moderate view, the international community has a collective duty to ensure that each state has the support that it needs to protect its own members against the standard threat of displacement. That assistance is a matter of justice, rather than charity. If it were a matter of charity, then members of the international community would be morally at liberty to supplement the protection of IDPs by their own states according to their discretion. Rather, I take the duties to assist in the protection of IDPs to be duties of justice, which their beneficiaries are owed by right. 13 Before unpacking and defending this view, however, it is worth contrasting it with two other possible views, which explain why I have called it the 'moderate' view.
First, consider what I will call the expansive view. 14 According to the expansive view, justice requires that the international community cancels out the effect of bad 'brute luck' (Dworkin, 2000: 73) in the distribution of the costs of addressing displacement between states. It says that as, a matter of justice, any burdens that states bear in protecting their members against the standard threat of displacement for which they are not responsible should be equalised. Adopting the expansive view, however, would require us to accept a radically altered conception of responsibility for protecting human rights in the international order. Its justification cannot stem from a reconstruction of the practices of the contemporary international order, and so I set it aside here.
Second, consider what I will call the restricted view, according to which international duties of justice, rather than charity, are grounded in responsibility for internal displacement. According to Laura Valentini (2013: 501), 'we have duties of justice towards the needy when our agency has contributed to their plight . . . and duties of charity when our 'hands are clean' but we can still help them at reasonable costs'. Valentini claims that this principle helps us to explain our moral intuition that those affected by the Haitian earthquake of 2010 are owed assistance as a matter of justice, while those affected by the earthquakes of 2011 in New Zealand and Japan are owed assistance only as a matter of charity. On Valentini's view, our duties of justice in the Haitian case arise from the fact that Haiti's vulnerability to the impact of natural disasters has roots in historic injustices, such as its colonial past and the reparations it was forced to pay to France, its occupation by the United States, and the economic liberalisation that was thrust upon it by the World Bank and the International Monetary Fund (IMF) as a condition for the receipt of loans (Valentini, 2013: 500).
As an interpretation of the practices of IDP protection in the international order, the restricted view is more plausible: the international legal framework around IDP protection does treat states' obligations largely as a matter of charity. Principle 25 of the Guiding Principles, for example, states that '[i]nternational humanitarian organizations and other appropriate actors have the right to offer their services in support of the internally displaced' and that '[c]onsent thereto shall not be arbitrarily withheld' (Deng, 1999: 492), which clearly does not constitute a firm obligation on the part of the international community to offer support. It is worth noting, though, that any firm legal obligation was resisted in the drafting process due to fears of intervention masked as 'humanitarian assistance', and that it was emphasised in the drafting process that 'if a government is incapable of providing protection and assistance, then the international community would be expected to act' (Cohen and Deng, 2016: 82).
Proponents of the restricted view are right to suggest that responsibility for displacement can ground duties of justice to assist in IDP protection. Where particular agents are responsible for displacement, they incur remedial duties of justice, which mean that they can be required to bear the costs of addressing displacement (at least, absent powerful countervailing reasons to redistribute those costs). 15 This explains why, for example, belligerents who cause displacement in wars, high-emitters who increase the risk of displacement through contributions to climate change, firms that cause displacement through extractive industries, such as oil and mining, and public authorities that cause displacement through large-scale development projects can all be held responsible to bear the costs of tackling the incidences of displacement that result from their actions.
The alternative that I propose, the moderate view, does not rule out that we may have duties of justice to assist in the protection of IDPs which stem from our agency. It is thus able to accommodate the claim that those who are responsible for particular instances of displacement are liable to bear the costs associated with that displacement. The moderate view, however, goes further than this in saying that the international community has duties of justice to assist in the protection of IDPs even in the absence of any duties arising from its members' agency. Even where no one is responsible for a particular instance of displacement, the international community has a duty of justice to assist states in protecting their members against displacement.
The defence of the moderate view begins by viewing the contemporary international order as, among other things, a regime of governance that distributes responsibilities to protect human rights between states (Hindess, 2003). In the first instance, states bear the responsibility for protecting the rights of their own members. As we saw in the first section, when that relationship breaks down, the individual is rendered a refugee and has a claim to international protection. In the case at hand, however, the relationship between the individual and the state remains intact, even if compromised, so each state is presumed responsible for protecting its members' rights. At first glance, this might lead us to think each state should be left alone to protect the rights of its own members.
Simply leaving each state free from interference to discharge its human rights-related duties, however, is insufficient to ensure that human rights are robustly protected, and expresses only a weak commitment to human rights norms. A genuine commitment to universal human rights protection in the international order should recognise that states need the 'positive' capacity to mobilise the resources necessary to address human rights threats (Ronzoni, 2012). Where states do not enjoy that positive capacity, there is a predictable risk that they will lose their standing as the guarantor of their members' human rights. Establishing provisions for assistance in protecting human rights can be a matter of preserving the 'background justice' of an international order structured by norms of state sovereignty and human rights protection (see Ronzoni, 2009: 242-249).
Addressing the standard threat of displacement requires states to mobilise the resources necessary to protect their members' rights. But states have different levels of capacity in this respect. Disadvantaged states may need to call on international assistance to protect their members against the standard threat of displacement, whether because of their own lack of capacity, because of the enduring impacts of historic injustices, or simply because accidents of geography leave them more vulnerable to displacement. They will foreseeably be unable to mobilise resources at a large scale in the way that addressing displacement requires. When acting alone, their immediate capacities to respond to displacement can be overstretched. As Roberta Cohen points out, situations of displacement are often particularly acute: Many governments do not have the resources, capacity or will to address the needs of the displaced, so that attention understandably shifts to the international community. The Guiding Principles on Internal Displacement [. . .] make clear that the international community has an important role to play in addressing the protection and assistance needs of IDPs, even though primary responsibility rests with their governments. Cohen (2007: 371) In such situations, the international community's response has important implications for the overall human rights regime within which the duty to protect against the standard threat of displacement is embedded. Where disadvantaged states can depend on assistance from the international community, they can maintain their status as the guarantor of their members' rights, and the role of the international community remains limited to supplementing their protection. Where they are unable to depend on the support of the international community, their failure to protect their members' rights can create situations where they risk losing their standing as the guarantor of their members' human rights. Individual incidences of displacement may not be enough for states to lose their standing in this way. But protracted crises of displacement can create threats to internal stability, threats of wider human rights violations, and even threats of state collapse (see Phuong, 2005: 219-226). In such situations, an 'offence' against the member-state relationship can easily devolve into a 'betrayal' of that relationship. Protracted crises of displacement are, as Cohen and Deng (1998b) put it, a 'symptom of state dysfunction'.
The international community's collective duty to provide assistance is a duty to create a bulwark against such situations arising in the first place. Collectively, the international community has a duty to eliminate predictable risks that disadvantaged states will lose their standing as the guarantor of their members' human rights. Admittedly, this standard is a little fuzzy. There is likely to be disagreement about when such risks are 'predictable' or 'eliminated' in practice. But this nonetheless gives a principled account of the international duties of justice associated with IDP protection, which can serve as a critical tool for evaluating state practice.
The justification for this duty of justice stems from the background normative structure of the international order. Its status as a demand of justice is a consequence of the way in which the international order distributes duties to protect rights. Maintaining the structure of human rights-protection is an important way of reconciling the international community's commitments to both human rights and state sovereignty. When the international community fails to provide the assistance that protects the structure of human rights protection, it creates threats to the legitimacy of an international order based on these norms.

Conclusion
This article has articulated an account of justice in IDP protection by normatively reconstructing the practices of internal displacement governance. We have seen IDPs are those who face reactive displacement, but whose relationship with their state fundamentally remains intact. Non-alienage, we saw, is only derivatively relevant as a proxy for identifying the status of this member-state relationship. We have also seen that internal displacement governance should be concerned with specifically reactive displacement, because of the threat it represents to the basic human interest in being able to form and pursue life-plans. The duties that states owe to IDPs -to reduce the risks of displacement to tolerable levels, and to restore background conditions of stability through return, local integration or resettlement where displacement does occur -respond directly to this harm. Finally, we have seen that states' duties to protect their member's rights with respect to displacement are embedded within a broader structure of human rights protection which assigns each state the primary responsibility to protect its members' human rights. To preserve that structure of human rights protection, however, the international community has a collective duty of justice to assist disadvantaged states in protecting their own members' rights with respect to displacement.