Refugees and the limits of political philosophy

ABSTRACT Everyone has the right to seek and to enjoy in other countries asylum from persecution, but in practice many people seek asylum and do not find it. Where asylum is in short supply, it may seem obvious and reasonable for philosophers to ask whether we can identify principles for prioritizing the asylum claims of some over those of others. In this paper I consider what kind of question this is, and whether it is one that philosophers are in a good position to address. I argue that philosophers have a number of powerful reasons to approach it with serious caution, and even to avoid it altogether. I outline some potential pitfalls of answering it, including the risk of normalizing violations of the principle of non-refoulement.

something (and something relevant) to say about it. As Joseph Carens noted over twenty years ago, it is one thing to ask definitional questions about refugeehood, or foundational questions about whether states have a moral right to exclude immigrants, or whether there ought to be a human right to immigrate, butso the common refrain goesit is not immediately clear how the answers to those questions help to contribute to national and global conversations, and guide action, here and now (Carens 1996). Perhaps political philosophers should (also?) be asking more questions of the kind that directly address today's political problems, and seek to contribute to political discussions and decision-making, by factoring into our theorizing more features of the world as we find it.
One deeply regrettable feature of the political world we inhabit is that, while everyone supposedly has a human right to seek and to enjoy in other countries asylum from persecution, in practice many people seek asylum and do not find it. As the organizers of a philosophy conference on 'refugees and minority rights' put it, we confront a situation in which various states 'are unlikely to accept as many refugees as they ought, and very few are likely to accept more than they are required'. In this 'non-ideal world', they asked, 'what are the right principles for admitting and rejecting refugees when asylum, whether permanent or temporary, is under-supplied … ?'. They also wondered whether it is even permissible (or prudent) to ask and/or answer questions about prioritizing between the claims of different refugees in the first place. 4 While I agree about the importance of philosophical reflection on pressing political issues, and share the readiness to participate in urgent debates, I have a number of concerns about engaging with the question of identifying principles for selecting between refugees when asylum is 'under-supplied'. The invitation to participate in the conference provided an occasion to think through these concerns. In the paper I consider what kind of question this is, and whether it is one that I am in a good position to address. For example, there are some questions with no appropriate answerperhaps each of the options is equally bad and all are impermissibleand in which context a philosopher's intervention may be (at best) redundant. There are some questions which call for the marshalling of various forms of empirical evidence or reference first-hand experience, in areas where a philosopher, qua philosopher, may have no special expertise. There are some questions which are difficult, contentious, sensitive, serious, and require nuanced responses. Here philosophers are likely to add caveats and provisos, but those details may be overlooked or ignored by the relevant audiences. There are some questions where a philosopher's answers could be used to lend justification to what seem like inexcusable policies, and so could contribute to harmful outcomes. The question about prioritizing between different refugees has elements of 'all of the above', and I will argue that philosophers have a number of powerful reasons to approach it, and questions like it, with serious caution. As a preliminary step, I think we need more conversations about the limits and potential pitfalls (as well as the possible benefits) of these interventions.
A number of the concerns I raise about philosophers engaging with this specific question will also apply in similar ways to a range of questions on other subjectsfor example in the 4 See the full conference outline here: https://philevents.org/event/show/62554. ethics of war, medical ethics, and animal ethicsand I will highlight some parallels during the course of the discussion. 5 I do not seek to argue that this case is sui generis, or that it is uniquely troubling for philosophers. 6 However, I do note that philosophical discussion of asylum and refugees at present is marked by a combination of the following features. First, it takes place against the backdrop of an existing international legal framework for refugee protection, at the centre of which are the 1951 Convention Relating to the Status of Refugees (Refugee Convention) and 1967 Protocol. This frameworkthough it remains fragile, undersupported, and vulnerable to non-compliancestill represents a major achievement of international cooperation. Philosophers may hope to extend its scope and robustness, encourage compliance with its norms, and promote greater solidarity and responsibility sharing among participating states, but we should be wary of any moves that might serve to destabilize and/or undermine it (the subject of Section 4). Second, we are writing at a time when refugee and asylum issues have risen to alarming levels of global political salience. Moreover, the rights of refugees are under scrutiny and attack across the worldtake the example of the USA, where President Donald Trump's administration is accused of waging a 'war on refugees', by erecting evermore barriers to people in search of asylum and dramatically reducing the number of places offered on its refugee resettlement programme (Holpuch 2019). Even though philosophers are not in control of (and often cannot anticipate) what will happen to their research once it is in the public domain, we all know that this is a profoundly sensitive moment for philosophical reflection on the ethics of migration, and that fact should give us pause. This kind of delicate timing, in the words of Jeremy Waldron, 'brings into play a certain duty of care on the part of the moral critic' (Waldron 2018, 93). 7 That, as Waldron emphasizes, is the price of 'relevance'. Third, seeking asylum is fraught with danger. Thousands of people die every year along migratory routes, others are imprisoned in detention centres, coercively separated from their families, subjected to violence and intimidation from law enforcement officers, and deported to countries in which their lives may be at risk. In short, the answers to these questions may have serious implications for the safety of real people. Fourth, philosophical work on this subject is often conducted from the perspective of people who are not themselves refugees and have limited experience of the challenges faced by people seeking asylum. It has a tendency to take on an 'us' and 'them' quality (even where that is seemingly 'benign', e.g. 'what should we do for them?'), as I explore in more detail in Section 7. The context in which we theorize matters, and the presence of these four background features underlines the importance of taking great care when approaching questions about refugees and asylum seeking.

Where to begin and what to leave in place?
'What are the right principles for admitting and rejecting refugees when asylum, whether permanent or temporary, is under-supplied in a non-ideal world?' There can be little doubt that we live in a non-ideal world. In political philosophy, one common way of understanding 'non-ideal' is in the Rawlsian sense of 'non-ideal theory'. Rawlsian ideal theory adopts the idealizing assumptions that (nearly all) the agents in question comply with the principles that apply to them, under 'realistic, though reasonably favourable, conditions'conditions amenable to the achievement and maintenance of reasonably just, democratic societies (Rawls 2001, 13). Rawls describes 'two kinds' of non-ideal theory: one which addresses noncompliance, and one which addresses unfavourable conditions (Rawls 1999, 90), but we could just as well describe these as two elements of non-ideal theory. According to Rawls, we should begin with ideal theory because it provides us with the fundamental blueprint, the reference point and the vision of what we can and ought to aim towards. It can also help us to work out our priorities, by indicating which current 'wrongs are more grievous and hence more urgent to correct' (Rawls 2001, 13). Much political philosophy, following Rawls, has taken the form of ideal theory.
Looking through a Rawlsian lens at the question of what to do when asylum, temporary or permanent, is under-supplied in a non-ideal world, we see there are at least two different features that suggest we are in the realm of non-ideal theory. First, the very existence of people seeking asylum is evidence that, among other things, this is a context of non-compliance and probably also that social and political conditions are less than favourable. The second element is that asylum, 'temporary or permanent', is under-supplied. These features are separable because we can at least imagine a context wherein the first feature was present but the second absentin which there were people in need of asylum in other countries, and yet asylum was available to all those in search of it.
There are various other ways of interpreting the 'non-ideal' part of the question. 8 We could understand 'non-ideal' here simply as 'the world that we have', as distinct from the one we (or I/you/they) might want. I will use non-ideal in that simple sense for now, so as not to be constrained within the terms of the ongoing methodological debates. The existence of people seeking asylum, and of more people seeking it than finding it, are certainly both features of the world that we have. My first concern is whether we should take these features as points of departure in our analyses, and consider which principles apply in under such circumstances, as the question about selecting and prioritizing between refugees may prompt us to do.
Any work in political philosophy begins with decisions about which aspects of the worldinstitutions, practices, beliefs, behaviours, relations, and so onthe author will take for granted for the purposes of analysis. For example, when a political philosopher asks whether states have a moral right to exclude immigrants, she is at least taking a world divided into states as a point of departure. She could choose to stick more closely to the world that we have by asking whether really existing states -Italy, Mexico, Thailand, and so onhave a moral right to exclude, or she could move further back and ask which kinds of states might have a right to exclude and under what kinds of conditions.
Where to begin is the pivotal question. There are all sorts of factors to consider and all sorts of reasons why a philosopher chooses to challenge some existing practices, institutions, assumptions and so on in a given piece of work, while leaving others in place from the start. Famously, Rawls described his own approach as 'realistically utopian', in that it 'extends what are ordinarily thought of as the limits of practical political possibility', while 'depict[ing] an achievable social world' (Rawls 1999, 6). So Rawls's ideal theory is removed from some key elements of the world that we have (such as widespread injustice), but his vision is constrained by what he thinks is realizable, given the 'laws and tendencies of the social world' (Rawls 2001, 4). Yet what is realizable? What count as the laws and tendencies of our world? What are the non-negotiables? What constitute the 'limits of the practically possible' (Rawls 1999, 12)?
Ideal, non-ideal, 'realistic', 'utopian', 'critical', or whatever else in between and beyond, political philosophers always begin with a series of decisions about which, how, and where (at what stage) parts of the world will feature in our theorizing; witness, for example, the significant and longstanding debate around the nature of 'facts' about the social and political world, and which, how, and where facts about the world should feature in political philosophy. 9 The question about principles for selecting between refugees where asylum is in limited supply asks us to factor in certain regrettable features of the world that we haveboth the existence of refugees, and the 'under-supply of asylum'and at an early stage of our analysis. What are the potential benefits of factoring in regrettable elements of the political world that we have into our theorizing and arguments at an early stageeven building them in from the outset? This is a contentious issue, and I can only treat it in a provisional and cursory way here. In brief, we might think that the likelihood of having some positive debate-informing and/or action-guiding effects (however those might be understood) increases as we theorize about contexts that are more recognizably like our own. 10 But a well-known and oft-mentioned concern about at least some versions of more 'worldly' theorizing is that it could beand often isexcessively conservative and concessive with respect to some of the unpalatable, unbearable, and unjust features of the world as we know it. 11 This can leave the impression that these features are simply inevitable, that there is little or no point in challenging or resisting them, and, crucially, might draw attention away from the deep injustices at work.

Temporary asylum
With that in mind, I will start with the question of when, if ever, we should take the under-supply of asylum, 'temporary or permanent', as a point of departure in our analysis. I will urge caution. Indeed, I will argue against offering principles for prioritizing between refugees where 'temporary asylum' is under-supplied.
First, what is 'temporary asylum' and in what sense could it be under-supplied? One plausible understanding of 'temporary asylum' is that asylum seekers who present themselves at or within a state's borders or find themselves under a state's jurisdiction are not turned away, rejected, or removed in any way that would be in violation of the state's 'non-refoulement' obligations. The non-refoulement principle, laid out in the 1951 Refugee Convention, ratified by 145 states, and also binding on the state signatories of the 1967 Protocol, is now considered a norm of customary international law and so binding on all states. 12 According to the principle as it is articulated in Article 33 (1) of the 1951 Convention, 'no Contracting State shall expel or return ("refouler") a refugee in any manner whatsoever to the frontiers of territories where his [or her] life or freedom would be threatened on account of his [or her] race, religion, nationality, membership of a particular social group or political opinion'. 13 'Temporary asylum' could also be understood in a more interim sense, as a middle ground somewhere between non-rejection prior to assessment of the asylum claim, and permanent residence status. So, for example, the state in which refugees have arrived might be willing to allow them to remain 'temporarily' within its borders during and beyond the period that their asylum claim is being assessed, while not offering the prospect of settling there. This interim version of temporary asylum is on a continuum with the non-rejection version. Given the prohibition against removing refugees to any place 'where their lives or freedom would be in danger on account of their race, religion, nationality, membership of a particular social group or political opinion', the state in which a refugee has sought protection cannot divest itself of its responsibility to that person unless and until some other suitable arrangement (such as voluntary repatriation, relocation, or resettlement elsewhere) has been found. In effect, then, an under-supply of 'temporary asylum' signals that the context is one in which states are not abiding by their non-refoulement obligations.
I propose that we should refrain from trying to identify principles for selecting between asylum seekers where any 'rejection' would be in breach of non-refoulement obligations. Instead, I would emphasize that when it comes to 'temporary asylum', there can be no countenancing of 'under-supply'. People seeking asylum must not be returned to their country of origin, or to 'any other place where a person has reason to fear threats to his or her life or freedom related to one or more of the grounds set out in the 1951 Convention, or from where he or she risks being sent to such a risk'. 14 As UNHCR (the UN Refugee Agency) explains, while the principle of non-refoulement 'does not, as such, entail a right of the individual to be granted asylum in a particular State', it means that 'where States are not prepared to grant asylum to persons who are seeking international protection on their territory, they must adopt a course that does not result in their removal, directly or indirectly, to a place where their lives or freedom would be in danger on account of their race, religion, nationality, membership of a particular social group or political opinion', and '[a]s a general rule, in order to give effect to their obligations under the 1951 Convention and/or 1967 Protocol, States will be required to grant individuals seeking international protection access to the territory and to fair and efficient asylum procedures'. 15 All the states supposedly participating in upholding the international refugee regime already accept this. As such, offering principles for selecting between people seeking asylum, at this stage in the process, would be worse than conservative and concessive. It would be a retrograde move. 16 That is not so much 'the world that we have' as 'the world that we had'. 13 UNCHR, 'Advisory Opinion on the Extraterritorial Application of Non-Refoulement Obligations', 2. 14 I note that there has been some legal debate about whether the duty of non-refoulement may be qualified in cases where that is necessary for the protection of states' vital national interests during situations of so-called 'mass influx'. See Hathaway 2005, 357;Costello and Foster 2016, 311. 15 UNCHR, 'Advisory Opinion on the Extraterritorial Application of Non-Refoulement Obligations', 3, emphasis added. 16 Thanks to Lea Ypi for pushing me to emphasize this point.
Here one might point out that, although many states have expressed their commitment to supporting the refugee law regime and thus to upholding the duty of non-refoulement, nonetheless violations of non-refoulement obligations are commonas when vessels carrying asylum seekers are intercepted at sea and 'pushed back' to the point of embarkation, or to places where they remain in danger. 17 We cannot ignore the fact of noncompliance with the requirements. The rule might be in place, but practice often does not line up behind it. In this context, is it not desirable to try to identify principles for selecting between people seeking asylum, even at the 'temporary asylum' stage?
In reply, a significant worry here is that offering up selection principles in this context risks normalizing deviations from the central pillarthe 'cardinal principle' 18of the refugee law regime. This kind of normalization of deviations over time can contribute to the erosion of the norm. As Waldron argues in the context of philosophical criticism of the laws of war, there is a risk that philosophical interventions could weaken what is already a vulnerable norm, 'without putting anything better or anything at all in its place' (Waldron 2018, 92). Rather than err on the side of legitimizing any departures from the (already weakening) consensus around the duty of non-refoulement, I maintain that it is better for political philosophers to repeat that the duty of non-refoulement should always be respected.

Permanent asylum
What about identifying selection principles where 'permanent asylum' is in short supply? Here the focus is on refugees obtaining permanent residence and being set on the path to citizenship in a state other than their country of origin. The state of permanent residence could be the country in which a refugee first sought protection, or could be a country to which they subsequently have moved, been relocated (e.g. under the European Union's Emergency Relocation mechanism, designed to share responsibility between Member States for the granting of protection to refugees arriving in Greece and Italy), 19 or resettled (in line with UNHCR's resettlement programme, which identifies refugees who will be transferred from the country in which they have sought protection to another country which is willing to grant them permanent residence and eventually citizenship) (UNCHR 2011). Which agents are or would be doing the selecting here? It could be an international agency, such as UNCHR; it could be the potential destination state; it could be the state of temporary protection; it could be a decision procedure involving refugees themselves, and so on.
Selection happens in practice, of course. For instance, UNHCR identified 1.19 million refugees as in need of resettlement in 2017, and was offered resettlement places for just 75,188 people. 20 How does UNHCR select between resettlement candidates? As they explain, they 'prioritize among possible cases by assessing the urgency of their individual resettlement need and the applicability of the resettlement categories in order to identify the cases to be submitted to a resettlement country' (UNHCR Resettlement Handbook 2011, 36 'Permanent asylum' has a different status from 'temporary asylum' vis-à-vis existing international legal norms. The duty of non-refoulement is a binding requirement on all states, and so I have argued that any under-supply of temporary asylum signals that at least some states have failed to fulfil their non-refoulement obligations. However, there currently exists no binding requirement for states to allow refugees to settle permanently within their borders or to offer relocation or resettlement places. So the concern about eroding an existing norm does not apply in the 'permanent asylum' case. Quite the opposite: there is the prospect of filling in a 'protection gap' in the refugee regime between non-refoulement and long term solutions (Goodwin-Gill 2014, 44). Furthermore, whereas concerns about 'fair shares' or 'accepting more than required' are inappropriate when the alternative to non-rejection risks violating non-refoulement obligations, as it stands the current system places the heaviest admissions and 'hosting' demands on those states (such as Turkey and Pakistan) bordering the countries from which refugees are fleeing. The protection gap also translates into a 'solidarity gap', as other states fail to cooperate and share in the responsibility of supporting the refugee regime. 21 When asking about principles for assigning permanent residence places, then, we might assume a backdrop of 'selecting' between refugees who already have some level of interim protection (better or worse, more or less adequate) in a state other than their country of origin or habitual residence. Where permanent residence places are undersupplied (presumably meaning that people are victims of protracted refugee situations and so cannot be repatriated, and have no prospect of long term solutions where they are), the people who miss out would be left with some kind of protection but in a state of limbo, without the right to remain permanently and build a new life in any country. Their situation could be more or less dire, depending on their personal and familial circumstances, the conditions in their country of interim protection, and so on.
The 'under-supply' of 'permanent asylum' may seem like an intractable problem, and that might incline us towards treating it as a starting point in our analyses. However, if we look more closely at current figures, we see that it could be a relatively simple problem to fix. All the world's current 25.9 million refugees (less than 0.4% of the global population) could be accommodated permanently, by states capable of doing so, if the resolve were there. 22 The much lower figure of 1.19 million refugees identified by UNHCR as in need of resettlement in 2017 could have been accommodated quite straightforwardly by just a handful of countries. That is not happening, because many states are not willing to act in ways that would facilitate it. To repeat, we have an international political system accustomed to tolerating the 'under-supply of asylum' from countries nominally supporting the refugee law regime. Therefore, while offering principles for selecting between refugees where permanent asylum is under-supplied does not carry quite the same risk of undermining a key principle of the international refugee regime, even so it seems sensible to fear that taking the under-supply of permanent asylum as a point of departure in our theorizing 21 See the 2018 Global Compact on Refugees, available here: https://www.unhcr.org/gcr/GCR_English.pdf: 'There is an urgent need for more equitable sharing of the burden and responsibility for hosting and supporting the world's refugees, while taking account of existing contributions and the differing capacities and resources among States. Refugees and host communities should not be left behind' (1). 22 UNHCR's figures in 2019, available here: https://www.unhcr.org/uk/figures-at-a-glance.html. Last checked August 2019.
is unduly conservative and concessive in the face of morally regrettable features of the world as it is now. It may give the false impression that practices and arrangements which could be different are in fact inevitable, while diverting attention away from the hypocrisy, non-compliance and injustices at play. This move is concerning to those of us who believe that a central role of political philosophy is critique, in some form. That does not necessarily mean that there is no place at all for such a question about prioritizing between candidates for permanent settlement, and for a studied response to it, but it does suggest that it is important to think very carefully about starting points.

Refugees
Does all this suggest that we should refuse to take the existence of refugees and people seeking asylum as a point of departure in our analyses? After all, where there are people seeking asylum, there are also agents who have failed to abide by other norms of international law and to discharge various duties. Why not simply insist that everyone should comply with all their human rights obligations and avoid any complicity in the creation and intensification of refugee situations?
It is possible to engage with issues relevant to asylum and refugees without explicitly engaging with the issue of asylum and the specific predicament of refugees. An argument in defence of open borders is a case in point. If state borders were more open to any comers, then refugees would encounter fewer obstacles on their paths in search of asylum. However, to address the situation of refugees and asylum-seeking directly, a theorist will have to engage with the fact of people in search asylum, and I think there are good reasons for taking the existence of refugees as a starting point in our analyses. We can do this without assuming that nothing can be done to reduce the number of people in search of asylum. We can address the issue(s) of why there are people seeking asylum, and can consider what various agents might do and/or ought to do to tackle the root causes that create and exacerbate refugee situations. For example, we might consider the role of the international arms trade in fuelling the conflicts which lead to mass displacement. 23 It is not too great a leap to imagine that some of the agents who figure in the causal story may be responsive to the right kind of reasons (or to the wrong kind of reasons).
At the same time, though, it does seem reasonable to treat the existence of at least some refugees as a likely feature of any world that we will have for the foreseeable future, and then to consider how to respond to the predicament of refugees. There are at least two important differences hereone practical and one ethicalbetween taking the existence of refugees as a starting point, and taking the under-supply of asylum as a starting point. 24 First, and more straightforwardly, to transition from where we are now to a point at which there were no people in need of asylum in other countries, we would have to come up with a solution to all the many and various factors that create and exacerbate refugee situations. That seems fanciful. Even if all the agents responsive to moral, legal and political arguments about human rights obligations and disposed to abiding by the relevant requirements actually complied with them, still some people will be the victims of civil and interstate wars, of climate-induced and post-conflict instability, and of widespread insecurity and disorder in the context of failed and failing states. 25 We have an international refugee regime built around the acknowledgement that states have primary responsibility for upholding the rights of their own members but that some states will not comply with their obligations. Some people will have to leave their countries in search of asylum elsewhere and, in those cases, other countries will have to assume international protection responsibilities.
Second, the 'under-supply' of asylum is in large part a product of non-compliance and unwillingness from states who are nominal participants in the international refugee protection regime and who are benefiting from the cooperation of others in that regime. Proposals about prioritizing between refugees are usually addressed to states (such as Australia, and various EU Member States) that rely on and demand compliance from other states, but who are unwilling to comply with the basic principles of the refugee protection regime when it does not suit them. In effect, we are saying to those states, 'you have agreed to participate in the refugee protection regime, you have accepted the obligations involved, you rely on the compliance of others, and yet you are not willing to comply now. In these circumstances, where you are refusing to comply with the requirements and so creating a problem of under-supply, here are principles for prioritizing between refugees.' In sum, a political philosophy enquiry which assumes, as a point of departure, that there will be some people seeking asylum may have the benefit of speaking to contemporary issues and contributing productively to public debates without (at least on the face of it?) appearing unreasonably concessive or conservative with respect to morally regrettable features of the world as it is now, but the same cannot be said in the case of the under-supply of asylum.

Isn't our input needed?
A political philosopher might share some of these concerns, but may also believe that it remains imperative to engage with questions about principles for admitting and rejecting refugees when asylum is under-supplied. We cannot ignore the reality that states are not providing sufficient opportunities for resettlement, for instance, and so we have to work with that fact. The input of political philosophers may help improve the quality of public debate on these matters, and/or may feed into political discussions and policymaking at some level and in some positive way.
I imagine this confidence in the potentially positive contribution of political philosophers on this topic comes from the sense that we are used to thinking carefully about questions that look a lot like this one in various respects, and so we can turn our attention to this issue using similar resources and thought processes. You might consider this to be one of a political philosopher's distinctive, transferable skills. Thinking about the principles which should guide decision-making is the bread and butter of moral and political philosophy (e.g. save the greater number? Distribute according to need/desert/strict equality?). Political philosophers are accustomed to questions like 'which are the correct principles for the distribution of x when x is scarce and demand outstrips supply?'. Isn't the question about principles for selecting between refugees just another version of that?
The problem is that we are not just engaging with interesting philosophical puzzles here. As I noted, if we are answering the question about the correct principles for admitting and rejecting refugees when asylum is under-supplied, with a view to having some impact on public debate and/or policy, then we have to bear in mind the predicament of those who lose outthose who are not prioritized according to the principles we have stipulated. These are not hypothetical examples. These are real people. Does our expertise extend to adjudicating between the claims of refugees?
It could, but not on the basis that we are political philosophers per se. Among other things, it would come from a familiarity with existing refugee resettlement programmes and the consequences that follow from the implementation of certain selection criteria (including, for example, the unintended consequences of those processes, such as the increase in clandestine and dangerous journeys in search of asylum). It would draw on the research of scholars in Refugee and Migration Studies (lawyers, economists, sociologists, etc), who have produced qualitative and quantitative studies about refugee movements, the practices of host countries, the effects of refugee-related policies, and so forth. It would mean engaging with refugees themselves, and listening to their testimony about what it is like to live in urban settings or camps, without permanent resident status. In short, it would mean acquiring knowledge of a range of empirical material which is emphatically not the bread and butter of moral and political philosophy. Why? Because how can we identify principles that are supposed to apply in this case unless we have a good understanding of the nature of the case? And acquiring the knowledge above is the minimum of what I think is required for a requisite understanding of the nature of the case. Since we have established the importance of the outcome for those on the losing side, we should not wade into this debate about real people in the real world without familiarizing ourselves with this kind of material. Now, you might think that we can get around this issue by suggesting a division of labour. Political philosophers identify the relevant principles, but it is people with expertise 'on the ground' who do what is necessary to feed these into public debate, lobby governments, develop policies, and work out exceptions to the rules, drawing on their superior understanding of what may or may not work in practice. But what role does that leave for political philosophers exactly? They could offer some general principles, such as 'adopt the decision procedure that causes the least harm' or 'prioritize those in greatest need of relocation', but those are not particularly helpful, and it is debatable whether practitioners really need us for those kind of generalities anyway.
Perhaps an area where the contribution of political philosophers is more apparent is not so much in, say, identifying principles for selecting between refugees when there are more people in need of resettlement than there are resettlement places on offer, but rather in identifying whether, when, and how states and other agents may prioritize their own particular interests and preferences when it comes to admitting or not admitting refugees. 26 Thinking about how to weigh competing interests, demands, and values in this way, and which should take precedence in cases of conflict, is very familiar to political philosophers. Practitioners may well have much to gain from engaging with political philosophers on questions such as the ethical significance of national communities and so on. Yet for a political philosopher to specify whether, when, and how states and other agents may prioritize their interests over those of refugees still requires them to have a good deal of working knowledge about the predicament of refugees, in the manner detailed above, because how could we make those sorts of judgements without an understanding of what this might mean for people in practice?
Interventions in the contemporary debate about the availability of asylum, without the requisite knowledge about the situation under discussion, could have negative effects on that debate by setting back public understanding of the issues, and could have negative action guiding effects more broadly, too. In short, this could do one thing that we all want to avoid: it could make things worse. Our arguments could be drawn upon to prop up harmful policies and reinforce divisive rhetoric, for instance.
You might be thinking that this gives too much credit to the potential for 'take-up' of political philosophy contributions to public debates. Are the relevant people really paying attention to what we write in academic journals and publish with university presses? Do our arguments have the prospect of informing discussion, behaviour, possibly even policy in this way? However, if you are not writing with the prospect of the ideas spreading beyond academic seminars, then why engage with these 'worldly' questions in the first place? Isn't the main rationale for taking the under-supply of asylum as a point of departure in our theorizing the prospect of guiding debate and possibly action (in some sense) in our world? 27 If that is the rationale, we can hardly downplay the potential for our work to have these effects while simultaneously making the case for doing this kind of work.
A possible response here is that what I have said so far reveals a certain kind of squeamishness about tough questions, an unwillingness to be drawn into messy and difficult debates with potential real world consequences, a retreat from the 'politics' part of political philosophy. One could maintain that we might not like it, but tough choices have to be made, and political philosophers should not shy away from making them. As David Miller writes, we might hope for a system in which all can obtain asylum when needed, 'but supposing this hope is unfounded: then it is better to say honestly that not everyone can be rescued, just as in the other cases where human rights are stakesuch as conflicts that require humanitarian interventionwe may have to acknowledge a gap between the rights of the vulnerable and the obligations of those who might protect them' (Miller 2016, 93). You might disagree with Miller on the substance here, but what about the sentiment that it is better to come out and say it?
In reply, political philosophers are not politicians. We do not have to be willing to speak on every topic, and we do not have to make life and death decisions as part of our job as researchers. We do not have to put ourselves into positions where we might be called upon to make terrible choices on behalf of others. It is (or at least it should be) up to us what we write about, and which questions we want to answer, and which parts of the world we will accept for the purposes of argument and which parts we will not accept. 28 Is it a dereliction of duty, in the sense that the consequences will be worse without our input? As I have suggested, UNHCR will still make decisions about selection criteria under 'non-ideal' circumstances, whether we participate in the discussion of selection principles or not. Are they making the 'wrong' decisions? As long as asylum is 'under-supplied' some people will lose out, and it will be terrible, and will not be much more or less terrible without our input. We may be at greater risk of making things worse by intervening without adequate care on this question than by steering clear of it. 29 Of course, instead of addressing this particular question, we could raise related questions about institution-building and obligations of responsibility sharing in the international refugee regime, about refugee-led proposals and solutions, and so on. This should not be read as a demand to stay out of these debates or away from 'nonideal theory' altogetherfar from it.
Aren't we perfectly positioned to contribute to this debate?
Up to this point I have been using 'we' to refer to 'political philosophers', which includes me. But this 'we' calls for further scrutiny. Who are 'we'? We are a diverse community of scholars. We come from a range of countries across the globe, we speak different languages, we are a variety of ages, religions, and so on. However, it appears that many of the people dominating the academic discussion do not have direct experience of being refugees, or of forced displacement, or of living in the countries that are home to over 90 per cent of the world's refugees. We should be aware that this can mean the absence of the voices and perspectives of numerous significant stakeholders, and an amplification of the voices and perspectives of a relatively privileged group. 30 There is a tendency for the 'we' to become a kind of shorthand for 'we who are not refugees; we who are not from the Global South; we who are citizens of countries relocating or resettling refugees; we who have obligations to "them"'. But this must not be a one-sided conversation, conducted from the position of just a few regions, standpoints, interests, and preoccupations. For instance, in recent years, much discussion in political philosophy on the subject of refugees has focused on questions about the responsibilities of statesoften so-called liberal democratic statestowards refugees, and in particular what constitutes a fair distribution of those responsibilities. Meanwhile, there is very little discussion of internal displacement, even though the global number of internally displaced people far exceeds the number of refugees. Why this focus? In all likelihood, it is because those are the pressing questions from where the most vocal 'we' are standing. Different sets of questions may feel far more pressing 28 Of course, in practice, this freedom is heavily constrained in a variety of social and political contexts. In addition, as Annamari Vitikainen points out to me, some might wonder about the extent to which it really is up to us what we write about and which questions we want to answer, given trends in funding regimes and the pressure to obtain large grants. Others worry about the silencing effects of social media 'pile-ons' and online 'mobbing', which 'discipline' academics and punish interventions on various issues. 29 For an argument that we, as political philosophers, should not necessarily answer all the questions we are asked, see Jubb and Kurtulmus 2012. 30 On the importance of engaging with and foregrounding the experiences of refugees, see Fine 2019. from other standpoints. 31 What is more, the answers to those questions may be received differently, too.

Conclusion: political philosophy with a clear conscience
My conclusion is that we do not need to answer questions about the principles for admitting refugees where permanent asylum is under-supplied, and we should avoid accepting the under-supply of temporary asylum as any kind of starting point. We could do more harm than good in answering these questions. And if we are going to try to answer them, then I think it is important to engage carefully with a range of expertise (beyond political philosophy) on the issue, while familiarizing ourselves with current policies and practices, and listening to a variety of stakeholders (beyond political philosophers). This is not a manifesto for retreating from philosophy, or from politics, or from the world that we have, but for approaching some issues and questions with cautionthinking about who we are, and what we can offer, while being mindful of our limits.