A liberal theory of asylum

Introduction

In recent years, western states have gone to great lengths to make their asylum systems less welcoming to refugee applicants. Australia for example triggered an international incident in 2001 when it forbade a Norwegian freighter carrying several hundred asylum-seekers from landing on its shores. In the USA, even before September 11, lawmakers so feared terrorists posing as refugees that they introduced an asylum policy that was so restrictive one observer has said it ‘essentially wipes out asylum as we know it’ (Fragomen, 1997: 443). The flip side of European Union members opening their borders to each other has been an increasing reliance on no-entry polices for asylum-seekers, to the point that refugee advocates now speak of Fortress Europe. Scholars who survey these and other developments have identified an international crisis of asylum. In the words of one analyst, ‘if the provision of protection for refugees is its central goal, then the system of asylum offered by Western states is currently in deep crisis. Over the last few decades, liberal democratic states have put in place barrier after barrier to prevent the arrival of rising numbers of refugees' (Gibney, 2004: 229).

This trend has contributed to a renewal of interest in Hannah Arendt’s diagnosis of the situation of refugees. Arendt drew a philosophical lesson from the refugee upheavals Europe experienced between 1914 and 1948. Then as now, refugees could not make appeal to rights they possessed in virtue of their citizenship, as they had lost the protection of their governments. The rights they invoked, rather, were the rights they possessed in virtue of being human. But as Arendt pointed out, European refugees routinely met with severe mistreatment. She pessimistically took this to show that human rights are illusory. Today, her argument is increasingly invoked by philosophers who also believe that the defining moral concept of our time has had its day, and that we need to go ‘beyond human rights', as one prominent neo-Arendtian puts it (Agamben, 2000: 15). ¹

Arendt’s argument should be of concern to anyone sympathetic to the plight of refugees. But her argument also poses a philosophical problem for liberalism. As we will see, Arendt thought the unreality of human rights was rooted in the fact that the earth is divided into sovereign states. Liberal theorists invariably accept some version of the state as a just institution. At the same time, they routinely endorse human rights. ² From an Arendtian point of view, this is a contradiction. Yet to date liberal philosophers have had little to say about the rights of refugees. To my knowledge, none have sought to defend liberalism from Arendt’s powerful challenge. ³

What follows is an attempt to outline a liberal theory of asylum. It begins by summarizing Arendt’s argument in a manner that emphasizes its ramifications for liberalism. Among the most important of those ramifications is the need for a model of asylum-seekers' rights that can be implemented in a world in which states exercise border control, whether or not border controls themselves are defensible at the level of ideal justice. The difficulty in enforcing rights for asylum-seekers in such a world is illustrated by reference to Germany and the crisis it experienced after introducing a constitutional right to asylum. I then go on to defend a model of asylum-seekers' rights that seeks to avoid the drawbacks of the right-to-asylum approach. Rather than commit liberal states to granting asylum to every genuine refugee who crosses their borders, this model would commit liberal states to recognizing asylum applicants as bearers of certain procedural rights. Crucially, the procedures in question can be carried out inside or outside the territory of the state in question. After outlining my theory, I conclude by considering some possible objections. For reasons I hope to make clear, a portable procedural model of asylum-seekers' rights ultimately represents liberalism’s best chance of upholding truly universal human rights.

Arendt’s case against human rights

Arendt’s attack on human rights occurs in The Origins of Totalitarianism, which contains a long chapter discussing the treatment of European refugees during the first half of the 20th century. The purpose of Arendt’s historical discussion is to show that human rights have not been effective in the case of refugees. In Switzerland, for example, Jewish refugees were violently turned back at the border and handed over to the Nazis. In countries where refugees were admitted entry, they were often unable to obtain legal work or residency papers, and so lived a precarious existence made up in equal parts of poverty, exploitation, and fear of deportation. This was the case even in states that had ringing declarations about human rights in their constitutions. France, for example, had been the birthplace of the rights of man, as human rights were first called, but it still treated refugees with severity, herding German Jews and others into disease-infested internment camps.

Arendt argued that the gap between the rhetoric and the reality of human rights was rooted in a tension between human rights and national sovereignty. Human rights, she noted, are supposed to be universal. Yet, we live in a world of particular states which do not affirm the rights of all human beings equally. Rather, they give priority to the rights of their own citizens. Nowhere is this clearer than in the realm of migration, where citizens enjoy a right of entry and other entitlements non-citizens do not. The result of this uneven distribution of rights was that when millions of desperate refugees arrived on their borders, liberal states exercised their sovereignty in such a way as to deny them admission. Governments supposedly committed to the rights of man did not enforce any rights beyond those of their own citizens.

Arendt took Europe’s refugee crises to show that human rights were ultimately illusory. The best refugees could hope for was to receive charity, as when receiving states occasionally allowed them to enter or obtain work. Arendt’s sceptical view of human rights is similar to that of Edmund Burke, and she concludes her discussion with a Burkean flourish. Her reflections on refugees, she writes, ‘confirm [Burke’s] assertion that human rights were an “abstraction,” that it was much wiser to rely on an “entitled inheritance” of rights which one transmits to one’s children like life itself, and to claim one’s rights to the “rights of an Englishman” rather that the inalienable rights of man’ (Arendt, 1967: 299).

Arendt’s reference to Burke should make clear that her quarrel was not with the concept of rights as such. Rights that we acquire in virtue of our membership in a polity she considered meaningful. But once someone had lost the rights that come with being a citizen of a state such as Britain, Arendt argued, that person was consigned to a position of rightlessness. This is the condition she took refugees to embody, a condition that ringing declarations of human rights did nothing to alleviate. In Arendt’s view, rights are only real if they are enforceable, and a refugee has lost the protection of the law of her country. A person in that situation has no meaningful rights to speak of, Arendt concluded, and so to keep invoking human rights is a sign of ‘hopeless idealism or fumbling feeble-minded hypocrisy’ (1967: 269).

This then is Arendt’s argument. In putting it forward, she sometimes gestures towards an alternative political arrangement that would avoid the problems she associates with states. ‘Human dignity needs a new guarantee which can be found only in a new political principle, in a new law on earth,’ she writes at one point, ‘whose validity this time must comprehend the whole of humanity while its power must remain strictly limited, rooted in and controlled by newly defined territorial entities' (Arendt, 1967: ix). Elsewhere in The Origins of Totalitarianism she also suggests that ‘a possible law above nations' could uphold the universal moral claims of humanity (Arendt, 1951: 436). ⁴

It is not quite clear how Arendt’s preferred alternative would improve on the present system of sovereign states. ⁵ Arendt’s references to an alternative arrangement are brief and lacking in detail. She never explains, for example, how new territorial entities would be different from states. Similarly, she does not indicate what type of entity would enforce a law above states or whether it would enjoy the prerogatives of sovereignty. If it did, it is not clear why it would not suffer from the same problems as states. Alternatively, if it lacked sovereignty, it is unclear how it would be able to influence the behaviour of states. For these reasons, Arendt’s cryptic references to a new system of international politics attract less attention than her well-developed critique of human rights. ⁶

That critique is one with special relevance for liberalism. There are contemporary political philosophers who reject human rights. One prominent example would be Alasdair MacIntyre, who has declared that human rights are ‘fictions... belief in them is one with belief in unicorns and witches' (1990: 68–9). MacIntyre’s position is unscathed by Arendt’s argument, as she targets a concept he does not wish to defend. Similarly, an anarchist has nothing to fear from Arendt’s claim that the illusory nature of rights is rooted in the existence of the state, as anarchists reject the state as a just institution. Unlike MacIntyre or anarchists, however, liberals have traditionally endorsed both human rights and the state. Arendt’s argument thus represents an immanent critique of liberalism, in that it challenges the compatibility of two of liberalism’s major commitments. ⁷

How might a liberal respond to Arendt’s critique? One possibility might be to endorse open borders. On this view, the fact that liberal states police their borders and turn many people away is a contingent rather than necessary feature of those states. Abolish immigration controls, this reply runs, and Arendt’s challenge disappears. For what she really highlighted is not an inevitable tension between national sovereignty and human rights, but one more reason not to stop anyone at the border, refugee or otherwise.

There are liberal theorists who endorse some version of open borders. ⁸ Such a view, however, has not traditionally been seen as an essential feature of liberal theory, and the open-borders reply to Arendt effectively makes the abolition of immigration controls a defining feature of liberalism. The open-borders position is also controversial. It is at least arguable that human beings need to live in communities that exercise some form of border control to maintain themselves. If so, then endorsing open borders may solve the problem Arendt highlighted only at the expense of creating other problems for liberal theory involving migration issues more broadly.

That is not the only problem with the open-borders reply. It also meets Arendt’s challenge on a high plateau of ideal theory. As Joseph Carens has pointed out, different moral questions reside at different degrees of abstraction from the world as it actually is. If we were to outline a completely ideal theory of justice, we would never mention asylum policy, as an ideal world would not contain the injustices that produce refugees. Once we take up asylum, however, we leave the purely ideal realm behind, as we are outlining an element of a theory of justice that applies when the world does contain refugees, and so fails to meet the standards of perfect justice. For this reason, Carens notes, ‘we should not assume that the just-world presupposition always offers a superior perspective on moral questions. Some of the most urgent moral questions simply disappear from view in a just world’ (2001: 20).

The open-borders reply to Arendt can admit the existence of refugees, and so is not a purely ideal theory. It nonetheless presupposes a high degree of abstraction from the international migration realm. Any response to Arendt that endorses a change to the status quo will of course contain some degree of abstraction from current reality, and this is true of the new asylum framework I propose below. The issue at hand, however, is one of degree. All western states currently exercise some form of border control, and there is no sign of them relinquishing this aspect of their sovereignty any time soon. A response to Arendt that stays entirely within an open-borders framework is therefore conditional on a sweeping change to immigration policy (far more sweeping than the one I propose) that may never actually happen. Even if the open-borders view turns out to be correct, therefore, it will still mark an important contribution to liberalism’s theoretical arsenal to put forward a model of asylum-seekers' rights that can be effective in a world of border enforcement, which is to say, in the world that we currently inhabit.

There is a second avenue by which liberals might seek to drain Arendt’s challenge of its force. It is to distinguish moral rights from legal rights. Arendt takes the historical mistreatment of refugees to show that human rights do not exist. But if we think of human rights as a moral concept, we are unlikely to concede that such rights cease to exist simply because they are often not upheld in practice. When there is some great rights violation such as ethnic cleansing or genocide, human rights advocates do not take the atrocity to show that the victims had no rights to speak of. Such atrocities show that the victims' rights were violated, but the concept of a right itself still has value. ⁹

This view contains an important truth. Often the correct response to human rights violations is to continue to affirm the value of rights. Such a response, however, is not an adequate response to Arendt’s critique of liberalism. The retreat to a purely moral understanding of human rights overlooks the fact that liberal theorizing about human rights has traditionally taken place against a backdrop of assumptions regarding what is politically possible.

We would not need a theory of human rights if the world always lived up to our standards of justice. Such a theory is, rather, meant to guide how things ought to be. In that sense, rights are inherently aspirational concepts. But they are not just aspirational. This can be seen by taking note of aspirations that would be out of place on a list of human rights. I might have an intense desire to speak to my dead relatives or to be able to leap over tall buildings in a single bound, but no liberal theory asserts such desires as rights. This is because rights are not only aspirations, but aspirations we expect to see realized. In addition, the realization of rights is regarded as a responsibility of states. Hence a desire to have a loving romantic partner is not normally regarded as a right. It is an achievable aspiration, but it falls outside the realm of goods we expect the state to guarantee, because the good in question is not one that can be delivered through state coercion, as love needs to be given voluntarily to be real. Normative theories of rights are thus dependent upon factual claims concerning what is not merely realizable, but what is realizable by states. John Rawls highlights this aspect of the liberal conception of rights when he notes that ‘fundamental principles of justice quite properly depend upon the natural facts about men in society’ (1971: 159). As Rawls goes on to elaborate, ‘principles of justice presuppose a certain theory of social institutions. Indeed one cannot avoid assumptions about general facts... If these assumptions are true and suitably general, everything is in order, for without these elements the whole scheme would be pointless and empty’ (1971: 160).

The vision of social institutions presupposed by liberalism’s account of human rights is that the state can be legitimately entrusted to enforce the rights of human beings seeking asylum. Arendt’s claim is that the liberal theory of rights is ‘pointless and empty’ for precisely the reason Rawls highlights: it is not consistent with the facts regarding the institution of the liberal state.

The open-borders and rights-are-purely-moral responses both evade, rather than address Arendt’s immanent critique. Both change the standards by which liberalism is to be judged, introducing new elements not already present in canonical theories of liberalism such as that of Rawls. The power of Arendt’s critique, however, is precisely that it judges the liberal account of human rights by the same standard employed by mainstream liberals such as Rawls in the passage above. A truly effective rebuttal to Arendt’s argument will thus itself be an immanent response. That is, a response that accepts (at least for the sake of argument) the underlying view of the relationship between a normative theory of rights and an institutional theory of the state found in both Arendt and Rawls. That shared view requires not merely outlining a theory of asylum-seekers' rights, but a theory of such rights that can be realized in a world of border enforcement, and enforced by sovereign states.

The rise and fall of a right to asylum

Arendt’s account suggests that citizen rights have a force and a reality human rights lack. One of the highest expressions civic rights can receive is to be enshrined in constitutional instruments, such as the American Bill of Rights or the Canadian Charter of Rights and Freedoms. This suggests that closing the gap between human rights and citizen rights will involve constitutional rights for refugees. A crucial question, however, is what form those rights should take. One approach would be to endorse a constitutional right to asylum. This approach, however, has already been tried in Germany, and it is instructive to note the problems that resulted.

In 1949, German lawmakers drew up a constitution for the new state of West Germany. They sought to make amends for the many refugees Germany had created under the Nazis. In order to do so, they included the following clause in West Germany’s constitution: ‘Persons persecuted on political grounds shall enjoy the right to asylum.’

West Germany received few refugees during the cold war, when the typical asylum-seeker was an Eastern Bloc defector. Following the collapse of the Berlin Wall, however, millions of Eastern Europeans could suddenly travel to a re-unified Germany. Germany has always maintained that it is not a country of immigration, and legally migrating to Germany is difficult for anyone not of German descent. The asylum system, however, was a possible alternative means of entry. This was true even for someone with a dubious claim to persecution, as the asylum clause meant he had to be admitted into the determination system, which could take years to decide his case. As a result of this and other incentives, almost 900,000 asylum claims were filed in Germany between 1990 and 1992, representing two-thirds of all asylum claims in EU states during the 1980s and 1990s (Gibney, 2004: 97). Overwhelmingly, the applicants were not fleeing persecution. Rather, they were economic migrants from Bulgaria, Rumania, and elsewhere seeking a better standard of living.

This provoked a crisis in Germany, one which included riots outside the German parliament and attacks on refugee shelters. Eventually, German lawmakers took the extraordinary step of amending the constitution and made their asylum system difficult to access. A key measure in this regard was the introduction of safe-third-country agreements. Such agreements stated that any asylum applicant who passed thorough an EU state before reaching Germany could be returned to that state. A new procedure was also introduced which saw claimants from countries with acceptable human rights records automatically rejected.

These procedures saw the number of asylum claims in Germany fall dramatically. In that sense, they solved the crisis. The new arrangement, however, has been criticized by refugee advocates. The revised asylum laws allow people to be automatically returned to countries with human rights records that are of debatable acceptability, including the Czech Republic, Romania, Poland, Ghana, and Senegal (Blay and Zimmerman, 1994). The network of safe-third-country agreements also makes it practically impossible to enter Germany and legally file an asylum claim. One estimate has suggested that 98 percent of pre-1993 cases would have been ineligible under the new rules (Bosswick, 2000: 51).

Obviously, more than one factor caused Germany’s crisis, including its descent-based model of citizenship. Important to note, however, is the role the asylum clause played. It was a magnet that attracted many asylum-seekers to Germany. Once large numbers of people began arriving, it only made things worse, as the increase in applications lengthened the time required to decide claims, which further increased the incentive for weak claims.

Arendt stressed that situations of mass influx represent the greatest possible challenge to refugee-rights enforcement. Germany’s case may be extreme, but it is not the only country to have faced an asylum influx. In recent years, Haitians and Bosnians have entered the USA and the EU in substantial numbers. Anyone who proposes a theory of asylum-seekers' rights therefore has to ask how his or her model would fare in a situation of mass influx, and how it would avoid the problems of the right-to-asylum approach.

A liberal model of asylum

I propose a rights-based approach to asylum that differs in two fundamental respects from the right-to-asylum model. Whereas West Germany enshrined the right to a particular outcome (that is, asylum), this model emphasizes the right to be subject to certain procedures during the refugee determination process. ¹⁰ Crucially, the procedures in question can be performed inside or outside the country that recognizes asylum-seekers as rights-bearing agents.

My model emphasizes the right to non-refoulement, which is the right not to be returned to a place of persecution. This is the fundamental right outlined in the 1951 Convention Relating to the Status of Refugees and as such, the moral foundation of international refugee law. ¹¹ A right to non-refoulement is conceptually distinct from a right to asylum. If someone from Sudan were recognized in the USA as a refugee, for example, and the US authorities transferred that individual to Canada, the USA would not have granted the individual asylum, but would nonetheless have respected her right to non-refoulement. When a state actualizes a right to asylum, by contrast, it actualizes a right that must be exercised within its own territory.

My proposal seeks to enforce the right to non-refoulement better by enforcing three procedural rights for asylum-seekers at a constitutional level. One is the right to an oral hearing to decide one’s case. Such a right is currently part of constitutional law in Canada, where it functions as a valuable, albeit limited safeguard against legitimate refugees being turned away without their claims at least being investigated. ¹² A second right involves representation by legal counsel. Studies of refugee applicants in the USA have found that applicants with counsel are four to six times more successful than those without (Government Accounting Office, 1987; Macklan, 2003: 23). Access to legal aid would reduce the possibility of legitimate claims being rejected because the applicant could not afford a lawyer. Finally, asylum applicants should enjoy the right to judicial review of detention decisions. In Australia, where mandatory detention of asylum applicants in prison-like conditions was widely used until 2005, numerous cases were reported of children and other asylum-seekers experiencing trauma and distress (Briskman et al., 2008: 111–214). ¹³ A right to judicial review would oblige governments to demonstrate some compelling need for detention, such as a legitimate security risk.

To see what is distinctive in the portable procedural approach, it is helpful to imagine different approaches to asylum inside a scalene triangle, which has angles of three different degrees. In the narrowest angle of the triangle, where there is the least room for human rights to be exercised, are clustered the asylum systems of most western countries. Here people seeking asylum have few if any constitutional rights enforced on their behalf. In the second angle is a German-style right to asylum, where rights have slightly more room to operate, at least in theory. In the third corner, with the most expansive room for rights to be exercised, is the portable procedural model. Its value is not that it is perfect or flawless (it is not), but that it has fewer shortcomings than either of the other alternatives our civilization has yet produced.

The core advantage of a portable procedural system over most existing systems is that it would enforce procedural rights for asylum-seekers at a constitutional level. One of the major problems with existing systems is that they place few if any limits on the policies politicians can introduce. This was noted by former British cabinet minister Richard Crossman. Looking back on a package of British immigration reforms of the 1960s, he observed that they contained ‘plans for legislation which we realized would have been declared unconstitutional in any country with a written constitution and a Supreme Court’ (Gibney, 2004: 117). Crossman’s observation explains why not merely the United Kingdom, but also Australia have been witness to some of the most extreme no-entry policies. Because both countries lack a national charter of rights, politicians do not have to contend with the possibility of a court challenge involving constitutional rights, and so have considerable latitude in how they treat refugees. ¹⁴ Even among countries that do have constitutional charters of rights, it is still the norm for the procedural and other rights they contain to be overwhelmingly rights of citizens rather than of non-citizens.

A portable procedural model of rights would be an improvement on most existing asylum systems because it would enforce three fundamental rights of asylum-seekers with the same powerful tool (that is, constitutional law) with which citizen rights are enforced. Rather than the exclusive prerogative of politicians, the justice of asylum policy would also be determined by courts. Were lawmakers to attempt to have asylum cases heard without respecting the three rights outlined above, they could be taken to court by the asylum-seekers in question or by human rights groups. Canada’s introduction of a right to an oral hearing was itself the result of a lawsuit brought forward by failed asylum claimants and two non-governmental organizations (NGOs). Canada’s subsequent no-entry policies, while all too real, have been less extreme than those of many other states, and this is arguably due to the well-established role of Canada’s Supreme Court in shaping asylum policy. Canadian lawmakers now have a strong incentive not to go too far in excluding refugees. Were they to do so, they could once again see rejected asylum claimants launch a successful constitutional challenge, overturning the government’s preferred asylum framework for being insufficiently rights respecting. The disincentive to violate asylum-seekers' rights is created not by actual constitutional challenges, but by their possibility, as Canadian politicians must draft asylum policy with an awareness that the Supreme Court is always looking over their shoulder. The portable procedural model would implant a similar awareness in the minds of politicians in other countries.

‘Constitutionalizing’ the procedural rights of asylum-seekers would not be a small change. It has often been observed that liberal democratic states currently seek to get around the obligations to refugees that international legal documents such as the Refugee Convention are supposed to impose on them. It is a slight misnomer, however, to identify states as the entities responsible for this trend. It is more accurate to say that politicians are the ones who seek to get around international law. When elected officials are given as wide a latitude as they currently enjoy in asylum policy, then asylum-seekers will inevitably receive poor treatment whenever it is in a leader’s or party’s interest not to recognize their moral claims. Unfortunately, admitting cultural or ethnic strangers is not always popular in democratic societies. The current trend of failing to live up to international norms is thus partly structural. Granting a greater role for the courts, which are less sensitive to the will of the populace, would improve the mechanisms of refugee-rights enforcement by reducing the manoeuvrability politicians currently enjoy in deciding how or even whether to implement minimum standards of justice for people seeking refuge.

This brings us to the second corner of the asylum triangle, that involving a German-style right to asylum. Germany’s experience confirms one premise of the portable procedural approach: constitutionalizing rights for asylum-seekers makes a difference in terms of what policies elected officials can introduce. As we saw, German lawmakers were forced to hear all asylum claims within Germany itself while the asylum clause was in effect. The long-term effect of the constitutional right to asylum was negative, however, ultimately contributing to the severe curtailment of the protection Germany offers to refugees. A crucial question therefore is what difference a portable procedural model would make in a situation of mass influx similar to the one Germany experienced.

We can answer this question by seeing how events might have gone differently had this model been in place in Germany during its crisis. Asylum-seekers who crossed into German territory would have been recognized as having a right to the three procedural safeguards mentioned above. However, the individuals would not have been entitled to those procedures being performed within Germany itself. The asylum-seekers could have been sent to a third country that agreed to uphold the three safeguards in question.

This arrangement would have two advantages over the right-to-asylum approach. The first is that politicians facing a mass influx would have greater flexibility to deal with the crisis than German lawmakers did. As we saw, Germany’s asylum clause obliged officials to admit every asylum applicant into Germany’s determination system. On the procedural model, by contrast, lawmakers could relocate asylum applicants to a sufficiently rights-respecting third country, and thereby break the vicious circle of ever-increasing unfounded claims and ever-lengthening determination times. By granting politicians a greater ability to defuse asylum crises as they are happening, the procedural model is less likely to create crises that can only be resolved by abolishing constitutional safeguards for refugees. A procedural model may therefore be better able to survive at least some situations of mass influx than the right to asylum proved to be.

The procedural model’s second advantage is that it would change the nature of third-country agreements. Such agreements are now a fact of life across the western world. That they result in refugee applicants being relocated from one liberal state to another is not in itself objectionable. What is of concern, however, is the circumstances under which such returns are made. Germany’s arrangement was typical in that its third-country agreements allowed it to adopt an ‘out of sight, out of mind’ attitude. If improper procedures were used or something went wrong after an asylum-seeker had been returned to Poland or elsewhere, it was of no concern to German officials. Germany’s responsibility for returned asylum-seekers ended once they left German territory. This approach risks seeing a refugee returned to a third country which in turn sends her back to a situation of persecution, a phenomenon known as chain refoulement. Even when this extreme outcome does not occur, returned refugees and asylum-seekers can suffer other negative effects, including becoming so-called ‘refugees in orbit’, shuffled from one country to another without ever having their claims heard.

The portable procedural model reduces the likelihood of such negative outcomes occurring by attaching conditions to third-country agreements. States that adopted the portable procedural model would be required to make sure that any third country to which they returned asylum-seekers would uphold the same procedural safeguards as the state from which they were being returned. Applied to the German case, an asylum-seeker would not lose the three rights outlined above upon being sent to Poland. Rather, those rights would ‘follow’ him or her over the border. Even if Poland did not normally supply asylum-seekers with legal representation, it would be obliged to do so in the case of returnees. Where Germany’s crisis was ultimately resolved by a severe reduction in refugee protection within Germany, a portable procedural model could potentially see future crises resolved by improving asylum procedures for at least some applicants in countries to which applicants were returned.

The portable procedural model involves a more powerful enforcement mechanism (constitutional law) than is currently brought to bear on asylum policy in most countries. In this way, it limits the power of elected representatives to employ substandard procedures during the refugee determination process. At the same time, however, it grants those same officials more flexibility in deciding where someone recognized as a refugee will eventually live, not necessarily forcing them to be admitted as residents in the first liberal state that recognizes their procedural rights. The portable procedural model thus avoids the extremes of the other two corners of the asylum triangle, one of which granted politicians too much freedom in how they treated asylum-seekers and the other of which genuinely limited what lawmakers could do, but in a manner that saw the rights of refugee claimants enforced at the expense of immigration control, and so was not viable over the long term.

The checks the portable procedural model would introduce against violating the rights of asylum-seekers would not come at the cost of abolishing border controls. It would allow asylum-seekers to be relocated to third countries that pledged to respect their procedural rights. It would also allow refugee-receiving countries to deny asylum-seekers work permits. The current international trend is for states to deny asylum-seekers the right to work even when there is no real need to do so. (Many governments seem not to realize that it is in their national interest to allow refugee applicants to work while their claims are being decided: when applicants have jobs they are less likely to seek welfare.) Nevertheless, withholding the right to work may occasionally be among the measures that policy-makers need to use to discourage false claims. Nothing in the portable procedural model rules it out.

A third border-control measure the portable approach allows is selective detention. The United Nations has stated that detaining asylum-seekers is ‘inherently undesirable’, particularly when asylum-seekers are housed with criminal offenders, a wrong that is only exacerbated when the asylum-seekers are children (Office of the United Nations High Commissioner for Refugees, 1999: 1). But as the UN also points out, there are cases in which adult detention can be justified. They include situations in which there is a need to verify a refugee’s identity; cases in which people have destroyed their travel documents; or to incarcerate a genuine terrorist or war criminal who has made an asylum claim. As is the case with withholding work permits, detention is currently overused by many countries, and often administered in a cruel and humiliating way. Yet given that detention can be justified in selective cases, it is worth noting its compatibility with the portable approach, on the crucial condition that detention decisions are subject to judicial review.

Fourth and finally, the portable procedural model would preserve the right of states to deport individuals who have been fairly found not to be refugees. Swift deportations for claimants rejected under reasonable procedures may be the most powerful disincentive against economic migrants lodging false asylum claims. By permitting this and other forms of immigration enforcement, the portable procedural approach would leave intact a country’s right to control its borders. It thus represents a better reconciliation of national sovereignty with human rights than our civilization has been able to manage to date.

I have stressed the practicalities of the portable procedural model in order to suggest how it might be implemented in a world in which immigration controls remain a reality. Such controls, however, give rise to normative concerns of their own. It is worth noting therefore that the portable procedural model does not oblige us to accept immigration control as a just practice. Even the staunchest global liberal could endorse the portable procedural model, for it does not challenge the case for open borders at the level of idealism at which that case is usually made. It, rather, seeks to offer an account of asylum-seeker’s rights one rung further down the idealism–realism scale, where refugee crises and border controls are both presupposed. The theory seeks to present a means by which the human rights of people seeking asylum can be respected, even in a world in which liberal democratic states without border controls remain an abstraction. In this way, the portable procedural model brings liberalism’s state-based theory of social institutions into greater harmony with its commitment to human rights. It seeks to demonstrate, contrary to what Arendt argued, that the two can in fact be reconciled.

In terms of the portable procedural model’s normative foundations, they are in one way similar to the current norm in the liberal democratic West and in one way different. The current norm is to see refugees as entitled to a right to non-refoulement rather than a right to asylum per se. The portable procedural model preserves this feature of the global status quo. Yet our political universe also sees the exercise of rights tied to territorial location. The portable procedural model makes the exercise of fundamental human rights less dependent on questions of location, without completely severing the link between rights and geography.

The fact that the portable procedural model makes the exercise of rights less dependent on geographic location marks its final aspect of superiority over the right to asylum. If we conceive of the world’s refugee population as being entitled to a right to asylum, a question arises as to where refugees enjoy this right. In particular, which country is obliged to grant them asylum? Such a question reveals the right to asylum to be more conditional and less ultimate than the right to non-refoulement. For there are states that have populations far smaller than the global population of refugees. In 2008, the global population of refugees was 9 million (not counting internally displaced persons). Iceland’s population is 317,000. If we concede that not all 9 million refugees have a right to move to Iceland, which would result in Iceland’s culture and institutions being overwhelmed, then we recognize a competing good that can in principle trump a right to asylum. The right to non-refoulement by contrast can be respected without obliging any particular state to admit refugees in numbers that would overwhelm local institutions. The claims of refugees and host societies can be simultaneously respected by transferring refugees from one rights-respecting state to another. Non-refoulement is thus a more ultimate principle, in the sense that its exercise is less conditional upon the circumstances of the society and state that actualizes such a right. The portable procedural model preserves a commitment to non-refoulement, but seeks to give it more powerful institutional expression.

The distinction between non-refoulement and asylum brings into view the relationship between rights and territory. In the modern world, whether a right is respected is currently contingent upon where the rights-holder is located. The institution that enforces rights is the state, and state’s have authority over particular territories. Currently, when lawmakers are deciding whether or how to respect the rights of refugees, that question overlaps with whether or not the refugee will remain within the same state’s borders. When states recognize an asylum-seeker as a genuine refugee, they commonly grant that same person residency, at least temporarily. The realization of the right of non-refoulement is thus manifest indirectly, through de facto asylum within the same state that first recognizes the asylum-seeker as a rights-bearing agent. The portable procedural model, by contrast, sees geography impose fewer restrictions on how the right to non-refoulement is upheld. Because the procedural and instrumental rights in question are portable, recognizing them is dependent not on granting refugees residency in any particular state, but in a certain type of state (that is, one that respects the three procedural safeguards in question, as well as the prohibition on non-refoulement). As such, a state can recognize the rights of arriving asylum-seekers without simultaneously making a commitment to residency. Decoupling a commitment to rights recognition from a decision in favour of residency increases the likelihood that asylum-seekers' rights will be recognized, rather than the current norm of violating them in order to keep out unwanted migrants.

A model of asylum-seekers' rights based on the principles I have outlined could be implemented in a variety of ways. In the Canadian legal challenge mentioned above, the Supreme Court of Canada found that the right to life, liberty, and security of the person upheld in the Canadian Charter of Rights and Freedoms entitles asylum-seekers to the right to an oral hearing. ¹⁵ Human rights jurists in other countries could potentially initiate similar litigation bringing their constitutional rights instruments to bear on asylum law. Less likely, but still possible is that when liberal states introduce new rights instruments or amend existing ones, procedural safeguards for refugees could be inserted. ¹⁶ In situations in which asylum-seekers have no legal protections at a constitutional level, the portable procedural model could inform the work of non-governmental organizations that lobby legislators on behalf of refugees. Non-governmental organizations could take the portable procedural model as one of their goals to work towards at the level of ordinary or administrative law. Portable procedural rights would serve as a standard by which to judge refugee-protection regimes which fall short of this model. Even where it is not legalized, popularizing it would put paid to the widespread view that controlling the border requires not even hearing the claims of desperate men and women seeking asylum. ¹⁷

Possible objections

Many objections could potentially be made to the framework of asylum-seekers' rights I have put forward. In the remaining space, I will address three possible criticisms. ¹⁸ The first criticism holds that the procedural model of asylum-seekers' rights does not address all of the possible ways in which states might violate asylum-seekers' rights. The second objection concerns the notion that refugees seeking asylum should be singled out as subjects of special concern, a practice Peter Singer has challenged. Finally, I will consider the view that the portable procedural model is an inadequate response to the Arendtian enforcement dilemma when national sovereignty is conceived of in popular rather than state-centred terms. The first criticism challenges the details of my particular model, while the latter two are more sweeping philosophical objections, concerning the practice of distinguishing refugees from other migrants and the conception of sovereignty my model relies on, respectively.

The framework of rights advanced here does not address all the ways states currently mistreat refugees. In the USA, for example, people seeking asylum are forbidden both from receiving welfare and seeking employment for six months after filing a refugee claim. This places them among the truly destitute (US Committee for Refugees, 1996: 6). Yet this and other forms of hardship are not addressed by the portable procedural model. In what sense, then, can it be justified in the name of human rights?

In response I would make two points. I have argued for a model of refugee rights to be enforced at the level of constitutional law. Constitutional rights place limits on the measures elected officials can implement. The experience of Germany, however, suggests that one cannot always anticipate the consequences of constitutional rights for non-citizens. I have not sought to constitutionalize additional rights beyond the three bedrock universal safeguards defended above in order to reduce the risk of creating a system with perverse incentives and unwanted consequences of the kind Germany experienced. It does not follow, however, that no other safeguards should be put in place for asylum-seekers. At the level of ordinary rather than constitutional law, they should enjoy a right to welfare, employment, and many other entitlements. Meeting these and other needs through ordinary law would allow politicians more flexibility should a circumstance of mass influx occur. Taking three core rights out of the hands of elected politicians in no way implies that refugees should be stripped of other legal rights at a non-constitutional level.

Some critics might then reply by asking what is so special about the three rights singled out above. Why should not a right to welfare be regarded as equally worthy of constitutional protection? Perhaps I am wrong to focus on only three rights. But if that is the case, the portable procedural model can be modified to take the overlooked right into account. It could be adapted so as to state that a refugee can be relocated to a third country where his or her right to welfare would also be respected alongside the three rights highlighted above. This type of criticism is thus not a rejection of the portable procedural model so much as a possible grounds on which to modify it. I would welcome its expansion to include as many rights as are feasible for it to contain. I have focused on the rights to an oral hearing, counsel, and judicial review partly because it is difficult to say in advance what all the necessary constitutional rights might be. But guaranteeing the three rights at hand is a necessary condition of extending justice to asylum-seekers, whether or not it is a sufficient one.

A second possible objection to the portable procedural model can be derived from the work of Peter Singer. He argues that while it is the norm for liberal states to treat refugees seeking asylum as subjects of special concern, the criteria on which they are distinguished from other people in need are dubious. According to the United Nations Convention Relating to the Status of Refugees, a refugee is someone with a well-founded fear of persecution for reasons of race, religion, nationality, membership of a particular social group, or political opinion. ¹⁹ Persecution, however, is a political phenomenon, and as Singer points out, people also migrate for economic reasons. ‘To distinguish between someone fleeing from political persecution and someone who flees from a land made uninhabitable by prolonged drought is difficult to justify when they are in equal need of a refuge. The UN definition, which would not classify the latter as a refugee, defines away the problem’ (Singer, 1993: 250).

Singer points to an additional problem with the institution of asylum. It grants weight to the principle of proximity (Singer, 1993: 254). Refugees who manage to enter a liberal state are entitled to the right not to be returned to a situation of danger in virtue of their location inside such as state. Refugees, however, are also huddled in camps in the developing world, and from the point of view of pure need, they make an equally strong, if not stronger, claim to assistance. The special obligation refugee-receiving states currently recognize in the case of refugees inside their borders is based not only on need, but also proximity. Singer is doubtful that such an approach is defensible. ‘It is difficult to see any sound moral justification for the view that distance, or community membership, makes a crucial difference to our obligations' (Singer, 1993: 232).

The theory I have defended here is not a theory of asylum in the narrow sense, in that it does not advocate that refugees always receive asylum in the first rights-respecting state they reach (my theory is a theory of asylum in a broader sense, in that it seeks to ensure refugees receive asylum somewhere, albeit not necessarily in the first liberal state they enter). Yet even though it differs from the conventional understanding of asylum Singer has in mind, my model does endorse the conventional distinction between refugees and economic migrants. Territorial proximity also plays a role in the portable procedural model, in that asylum-seekers have to be physically present inside a rights-respecting state in order for their procedural rights to be recognized and enforced. For these reasons, the portable procedural model needs to be defended in the face of Singer’s criticisms, notwithstanding its differences from asylum as we currently know it, and notwithstanding that Singer’s objection also applies to many other models of asylum law (including the other two corners of the asylum triangle described above).

In responding to Singer’s first criticism, it is helpful to distinguish between people facing different degrees of economic hardship. A refugee is someone who fears serious persecution. The economic equivalent of this is not someone who immigrates in the hope of obtaining a better job, but someone facing severe circumstance of the kind Singer mentions, such as drought or famine. A key question arises, however, when we ask what people in such dire economic circumstances really need. Consider Singer’s example of someone experiencing a water shortage. In countries facing famine or drought, the poorest of the poor do not have the resources to immigrate. Such individuals live in a state of absolute poverty, which precludes meeting their most basic needs (Singer, 1993: 219). Moreover, there are usually far better ways to assist them than to go through the expensive process of flying them to the West. As Matthew Gibney puts it, ‘in the case of victims of famine or natural disasters, it is easier for outside parties to deal with the threats people face by exporting assistance or protection (food, building supplies, clean water) to people where they are than to arrange access to asylum’ (2004: 8).

Gibney’s observation highlights the basic problem with Singer’s economic objection. It does not adequately separate moral status from legal status. On a moral level Singer is entirely correct: people in severe economic circumstances make just as strong a claim on us as do refugees fleeing persecution. The special status refugees enjoy, however, is legal rather than moral. They are entitled to priority in the immigration queue because unlike people facing economic deprivation, there is no other way to address their most urgent material need. Insisting on this legal priority is compatible with recognizing the equally serious moral claims of people facing severe poverty. Such claims, however, are most effectively dealt with outside the context of migration law.

Turning to Singer’s proximity objection, he is again correct at a moral level. Refugees in overseas camps are just as deserving of our concern and aid as are those who manage to reach our borders. Singer’s argument, however, should be taken to show that we need to do more for refugees in camps, rather than lessen our commitment to asylum. People fleeing persecution have arrived in liberal states in large numbers for many years and will continue to do so as long as persecution exists. It is only realistic to have in place a legal regime that seeks to avoid returning them to danger, regardless of what framework we put in place for other people in need, including other refugees.

In terms of meeting the needs of refugees in overseas camps, there is no question that liberal states could do far more for them than is currently the norm. As Singer notes, we do less than we should for the distant needy of any kind, whether or not they are refugees, and instead devote considerable resources to consuming luxuries we do not really need (Singer, 1993: 221). But given how little western states spend on all types of foreign aid, it seems unlikely that domestic asylum policies are a special barrier preventing overseas refugees from receiving more assistance. One can also imagine a world in which refugees both languish in overseas camps and are invariably returned to danger every time they reached a liberal state. Such an arrangement would do nothing for camp refugees, whose needs would remain equally pressing. On both a causal and conceptual level, then, recognizing the needs of asylum-seekers in no way prevents us from doing more to live up to our moral obligations to refugees overseas.

Indeed, there is some evidence that on a political level, recognizing the rights of asylum-seekers is more conducive to meeting the needs of refugees in camps than is failing to recognize asylum-seekers' rights. Refugee camps are often located in a state neighbouring a crisis zone. If refugees in the developing world cannot at least flee their country of origin, they will be even worse off than they are in a camp. In recent years, however, there have been cases in which industrialized states have made their asylum policies more restrictive, only to see refugee-receiving states in the developing world follow suit and introduce even more callous measures of their own.

In the early 1990s, for example, the USA introduced a policy of interdicting Haitians at sea and returning them to Port au Prince without even the possibility of filing an asylum claim. ²⁰ This severe and unprecedented policy was cited by the government of Thailand in 1992 when it engaged in mass expulsions of thousands of Burmese and Cambodian refugees (Guest, 1995: 80). An even more dramatic shift occurred in Tanzania, traditionally one of Africa’s most generous refugee-receiving nations. When 40,000 refugees tried to enter Tanzania from Burundi in March 1995, the Tanzanian government closed its border to prevent 20,000 of them from entering and announced plans to expel all refugees already inside its borders. Speaking at a refugee conference six months after the border closure, Tanzania’s foreign minister singled out the US interdiction programme as a precedent that had emboldened his government. ‘Citing the example of the Haitian refugees', a conference organizer noted, ‘[the minister] said that it was a double standard to expect weaker countries to live up to their humanitarian obligations when major powers did not do so whenever their own national rights and interests were at stake’ (Rutinwa, 1999: 20). ²¹

Thailand and Tanzania are only two examples of a global trend. In the words of an analyst with the United Nations High Commissioner for Refugees, ‘nations that absorb the most refugees in Africa will often cite the EU or U.S. tightening their policies as a rationale for them to tighten their own policies' (MacDonald, 2004: 13). In dealing with the situation of refugees seeking asylum, we are thus faced not only with their needs, but also with the question of what kind of example to set to countries which house large numbers of refugees in camps. In so far as we want to improve the situation of overseas refugees, we should hope that liberal nations set an example as model states of first asylum. This will decrease the possibility that governments in the developing world will place overseas refugees in even worse circumstances than they occupy now.

This brings us to the third and final objection. It challenges my proposal at an even deeper level than the criticism derived from Singer, and does so by calling into question the understanding of sovereignty I have taken for granted.

Arendt scholars have long noted an ambiguity in Arendt’s use of the term ‘national sovereignty’. ‘For Arendt, this refers to two separate principles, although she does not always clearly distinguish between them,’ Bridget Cotter has pointed out (2005: 97). State sovereignty is associated with the post-Westphalian international system, which treats states as ultimate legal authorities within their borders and only within those borders. Popular sovereignty by contrast refers to the right of a people or polity to exercise self-determination, most obviously in areas such as electing leaders or deciding which outsiders are allowed to join the political community. That these two forms of sovereignty do not refer to the same thing is evident in the fact that they can potentially come into conflict, as when a polity seeks to participate in an election which an undemocratic government forbids.

As Cotter and other interpreters of Arendt have noted, Arendt was quite critical of both state sovereignty and popular sovereignty. A key question, however, concerns which of the two can be said to be the final source of the problem she identified regarding human rights. Most often Arendt suggests it is states, as when she speaks of ‘the very institution of a state, whose supreme task was to protect and guarantee man his rights as man’ (1967: 230). There are moments, however, when she appears to have polities in mind, as when she characterizes the rights of man as the view that rights ‘should be guaranteed by humanity itself’, and goes on to express doubt as to whether this would be possible even under a world government, given that the human polity is capable of liquidating minorities (Arendt, 1967: 298).

By proposing constitutional safeguards that would rule out certain government actions, my focus up to now has been on state sovereignty. I have taken such an approach because it seems in keeping with the argument Arendt makes most of the time. ²² It is worth asking, however, whether the portable procedural model can serve as a safeguard against human rights violations caused by exercises of popular rather than state sovereignty.

The sovereign state first arose in Western Europe in the 13th or 14th century, when England and France became kingdoms with settled borders. ²² To see the difficulty of safeguarding the rights of asylum-seekers against exercises of popular sovereignty, it is helpful to image a hypothetical scenario taking place thousands of years ago, before any states existed. Let us imagine that in such a state-free world, a group of people are travelling by boat and come across someone treading water in distress. The person in the water desperately wants to be taken on board. The people in the boat enjoy popular sovereignty. They can determine who joins their group, which in this case involves being admitted onto their vessel. Seeking to enforce the rights of asylum-seekers against polities truly determined to exclude them is akin to asking if there is a rule that can be implemented on the boat that would make its operators pick up the person in the water, even when everyone on board was staunchly opposed to doing so.

In both cases, the answer is no. It is no more possible to implement an internally enforced rights mechanism that will guarantee the rights of outsiders against a polity universally opposed to admitting them than it is possible to implement a rule on the boat that will make its unwilling pilots stop for the person in distress. Although there are many differences between ancient seafarers and modern polities, in both cases the underlying problem is the same. In both cases, we have conceived of the group in question as not being subject to any external authority. In both cases, therefore, a rule to admit outsiders can only be enforced by the same group determined to violate the rule at hand. But there is no rule that we can expect to be upheld when its enforcement is left up to the very people that give rise to the need for the rule in the first place. Either sovereignty can be exercised by a group determined to exclude or such a group can have its sovereignty compromised by having an admission rule enforced from outside. But both outcomes cannot obtain at once. The portable procedural model is no more able to get around this problem than is any other legal mechanism enforced by an autonomous polity.

Conceding this point, however, does not entail rejecting human rights. If constitutional rights instruments cannot force a hostile and truly sovereign polity to respect the moral claims of outsiders, such rights instruments are equally ineffective in upholding the rights of insiders in similar circumstances. Applied to the boat scenario, the same problem would apply if the issue in question concerned whether the people on board were going to throw one of their members overboard. (We might wonder if an especially small group was capable of doing so, but our discussion concerns cases in which their ability to exclude someone is not in question: this is part of what it means to be sovereign.) Arendt’s argument was meant to highlight a special problem unique to the rights of man rather than citizens. The boat scenario, however, highlights a theoretical problem for any theory of rights, whether it is organized around the principle of humanity, citizenship, or any other concept.

This is because the problem at hand will exist where any group of human beings enjoys popular sovereignty and is determined to exclude outsiders. This is the case whether the group in questions inhabits a territorial state or a world government. It would also be the case if the people in question inhabited one of the many historical entities that, unlike states, do not assign authority to a set territory with fixed borders, such as an empire, city-league, feudal network, nomadic band, or anarchist polity without political structures of any kind. ²⁴ This is because popular sovereignty is ultimately rooted in human plurality rather than any particular political institution. An inability to restrain a powerful public determined to exclude is thus not a special weakness of human rights in a world of sovereign states, but a problem no theory of justice can rule out in any political universe in which one or more polities are truly sovereign.

From the perspective of our political universe, the problem of a sovereign public united in a project of exclusion is one that resides at a high level of abstraction. The standard we have been concerned with is whether laws can prevent human rights violations in situations in which entire polities are committed to violating the rights in question. A rights regime, however, can fail to meet this daunting standard yet still have value. The civic rights Arendt considers it wise to rely on, for example, have often failed historically to ensure justice for all citizens, yet she does not take this as grounds to reject the very notion of civic rights. Similarly, although the portable procedural model may not serve as an absolute guarantee in the case of universally hostile polities, or states in which there is no will to respect human rights, this is not grounds to reject it as worthless. In modern liberal states, it is rare to encounter entire polities that are monolithically committed to exclusion. ²⁵ More common is to find a range of views within a given polity, with some portions of the populations committed to excluding refugees, others favouring inclusion, and still others oblivious or indifferent. It is equally common for modern polities to be influenced at least to some degree by outside entities such as NGOs or the United Nations. Against the backdrop of a plural and divided polity in dialogue with outside forces, which is the typical polity we find in liberal states, constitutional law remains a powerful enforcement mechanism. That constitutional law may not be sufficient to uphold rights in all possible worlds does not call into question its power as an enforcement mechanism in this world. For this reason, switching our focus to popular sovereignty does not call into question the value of the portable procedural model. Although such a model cannot prevent every form of exclusion the human condition gives rise to, it will solve many of the problems asylum-seekers currently face, which is sufficient testament to its worth.

Conclusion

If the portable procedural approach were adopted, it would make the human rights to which asylum-seekers now appeal more similar to citizen rights. Arendt thought it was impossible to bridge the gap between these two understandings of rights in a world of sovereign states. The model of asylum-seekers' rights outlined here, however, does not call a state’s prerogative to control its borders into question. Rather, it seeks to extend to asylum-seekers three procedural entitlements that we take for granted when it comes to the rights of citizens, who enjoy all three safeguards defended here when they are accused of a crime. Were the portable procedural model adopted, I believe it would represent our civilization’s best change of reconciling the existence of sovereign states with the aspiration of human rights. That no state currently follows this model, not to mention the current treatment of refugees, should remind us how far our world is from justice for non-citizens. Two centuries after the Declaration of the Rights of Man, human rights is still a radical creed.