Introduction

Migration law and the inclusion of noncitizens are often framed as a matter of nation-centered concepts of justice. But this framing is often rooted in concerns about the fair treatment of noncitizens who have become part of a national community while remaining irregular migrants without lawful status. In the United States, for example, these concerns have prompted the emergence of a civil rights framework as its version of a nation-centered approach to migration and the rights of migrants (Motomura, Reference Motomura2020: 460–479).

In recent years, two trends have cast doubt on whether a civil rights framework or any nation-centered justice system for assessing migration and the rights of migrants can be effective or complete. One trend is the externalization of borders and border controls. The second trend is the shift in political focus away from irregular migrants inside a country and toward newcomers fleeing international conflicts, civil wars, environmental degradation, poverty, and famine. These two trends have combined to intensify attention on the international system for protecting “refugees.”

1 Refugee Protection

If migrants win recognition as refugees in destination countries, they gain favorable treatment under international and domestic law – typically a grant of asylum, durable residence, and access to citizenship. This widely accepted refugee exceptionalism (Carens, Reference Carens2013: 194–199; Miller, Reference Miller2016: 78; Walzer, Reference Walzer1983: 48–51) is the fragile core of a system that offers little or no protection to many of today’s migrants, who leave their homes owing to dire conditions but are outside the legal definition of “refugee” (Hamlin, Reference Hamlin2021: 1–9; Keyes, Reference Keyes2017: 138–147).

This system arose in modern form soon after World War II. Many countries had turned away people who later perished in the Holocaust. These countries’ responses led to the dominant refugee protection paradigm, including the foundational 1951 Geneva Convention Relating to the Status of Refugees. Its basic guarantee is nonrefoulement – the duty to not return people to persecution on account of nationality, race, religion, political opinion, or membership in a particular social group.Footnote ¹

This protection scheme reflected its origins in the chaos and suffering of postwar Europe, at the end of a cataclysmic conflict, and also as the continuation of geopolitical chasms that would persist much longer (Judt, Reference Judt2005: 28–31; Long, Reference Long2015: 4–5). The 1951 Convention was originally limited to migrants displaced by “events occurring in Europe before 1 January 1951.”Footnote ² Signatory countries obligated themselves to protect refugees, but they remained free to refuse “economic” migrants (Karatani, Reference Karatani2005: 541; Long, Reference Long2013: 13–21). Predating much of modern human rights law, refugee protection emerged as an exception, not a challenge, to sovereign control of national borders (Chetail, Reference Chetail and Rubio-Marín2014: 23–24, 39–40). Firmly rooted in Cold War politics, the system gave the United States and countries in Western Europe the latitude to recognize anyone who managed to flee the Soviet Union or its satellites as a refugee from Communism.

This legal structure is commonplace in the Global North, where countries brought refugee protection into domestic law in two types of schemes, both distinct from other migration regulation. An example of the first type is the US Refugee Admissions Program, which admits refugees from outside the country. The US president consults with Congress before setting an annual limit, subdivided among regions of the world. This number is much lower than the millions of people worldwide who might qualify, so selection criteria – including region, degree of threat, and US ties – are strict. Once admitted, refugees routinely become permanent residents, and many naturalize as citizens.

The second type of protection adopted into domestic law consists of asylum and related forms of protection that are available only at or inside the national border. In the United States, if an applicant makes the required showing of persecution on account of enumerated grounds, government officials may exercise their discretion to grant asylum. Unlike refugee admissions, asylum grants are not limited in number. But like refugee admissions, asylum routinely leads individuals and their spouses and children to both permanent residence and citizenship. Some persons not granted asylum may still be protected by withholding of removal or the Convention Against Torture from return to a country where they face peril.Footnote ³ These protections have requirements and benefits that differ from asylum.

Applicants for both refugee admissions and asylum in the United States must meet the same statutory definition of refugee, which requires showing “persecution or a well-founded fear of persecution on account of race, religion, nationality, membership in a particular social group, or political opinion.”Footnote ⁴ But refugee admissions differ from asylum in both practice and politics. Refugee admissions are extremely selective. Until recently, the annual limit fluctuated between 70,000 and 100,000 (United Nations High Commissioner for Refugees [UNHCR], 2019). The Trump administration slashed it to 45,000 for 2018, then 30,000 for 2019, then 18,000 for 2020.Footnote ⁵ President Biden returned to prior practice by announcing an annual limit of 125,000 for 2022 and 2023,Footnote ⁶ but actual admissions have lagged under these limits in recent years (Migration Policy Institute, 2022).

A key political difference between overseas refugee programs and asylum is that disappointed applicants for refugee admission are far away and have little recourse, so governments retain substantial control. For asylum, however, governments must work much harder if they want to control who reaches their borders and applies.

2 Exceptionalism under Pressure

The political sustainability of refugee exceptionalism is delicate. Enough people must see the number of asylum applicants as low enough to coexist with selective admissions for other migrants. This is not a matter of absolute numbers. Politics, including racial and religious perceptions, can fuel intense hostility toward asylum seekers from unfamiliar lands, even if few migrants actually arrive.

Domestic political viability can require that protection be viewed as an exceptional act of sovereign grace for extraordinary reasons, not as a matter of right that could undermine control over national borders (Aleinikoff & Zamore, Reference Aleinikoff and Zamore2019: 16; Martin, Reference Martin1990: 1266–1270). As long as asylum is seen as exceptional, few will associate it with legalization or amnesty, even though asylum likewise gives lawful status to migrants who might be barred or expelled because they lack permission to stay (Motomura, Reference Motomura2014: 195–196).

The second precondition for refugee exceptionalism to be sustainable politically is confidence that “refugee” can be defined consistently and fairly. Protection is not supposed to be for voluntary migrants – thus not for “economic migrants” – but only for forced migrants, and then only for some forced migrants (McAdam, Reference McAdam2012: 98).Footnote ⁷ This line-drawing generally limits refugee protection to migrants forced to move by specific events, not by the cumulative effects of deteriorating conditions – such as climate change – that magnify the consequences of political dysfunction (Lister, Reference Lister2014: 618).

Events have undermined these preconditions. Most forced migrants must make harrowing journeys before they reach distant places of refuge.Footnote ⁸ This fact has tempted the Global North to assume that countries closer to migrants’ homes would host them indefinitely. But geographic insulation is unreliable. On the US–Mexico border, for example, the number of migrants arriving from El Salvador, Guatemala, and Honduras has increased dramatically since the mid-2010s (Fernandez & Ferman, Reference Fernandez and Ferman2019).

To be sure, the overall number of Central American migrants coming to the United States remains a small fraction of new arrivals. Surreptitious border crossings are much less frequent, and the US unauthorized population has declined (Bolter & Meissner, Reference Bolter and Meissner2018). In Europe, migrants from the Middle East are far less numerous than the highs in 2015 (Nasr, Reference Nasr2019). But perceptions of threats to border security have been persistent in news cycles, and asylum seekers remain central to politics in destination countries (Kingsley, Reference Kingsley2018). As more migrants arrive, politicians and media can make refugee “crises” loom large, even if what seems like unprecedented migration is not all that new, and is in fact long familiar in the Global South. As one commentator wrote, “the populist narrative on migration is number-proof” (Dalhuisen, Reference Dalhuisen2019). Invented notions of crisis can amplify narratives of asylum claims as fabricated and asylum seekers as criminals, leading to demands for zero-tolerance responses (Zengerle, Reference Zengerle2019).

3 Managing Refugee Protection

One response by governments to perceptions of unprecedented large-scale migration has been to limit the number of asylum seekers, taking advantage of the Refugee Convention’s near-silence on how protection is implemented (FitzGerald, Reference FitzGerald2019: 41–251). One strategy shifts borders outward (Benhabib, Reference Benhabib2020; Shachar, Reference Shachar2020b). The US Coast Guard has long kept Haitians and many Cubans away from US shores.Footnote ⁹ In the Mediterranean, the dominant strategy is to interdict migrants on boats before they reach Europe (Pianigiani, Horowitz, & Minder, Reference Pianigiani, Horowitz and Minder2018). Australia combines interdiction with long-term detention of asylum seekers in camps in Papua New Guinea and Nauru (Polakow-Suransky, Reference Polakow-Suransky2017).

Governments adopt other strategies of migration control outside physical borders and on the soil of other countries (Zolberg, Reference Zolberg2006: 110–113; 264–267). The United States enlists Mexico’s help to impede the passage of Central American migrants to the United States. A 2016 agreement between the European Union and Turkey largely blocks Middle Eastern migrants from reaching Greece and other EU countries (Collett, Reference Collett2016; European Council, 2015, 2016). Several EU countries have arranged with Libya, Niger, and other African countries to limit northward migration (Hooper, Reference Hooper2017) (European Union, 2018). Similar in effect are requirements that asylum seekers apply in the first “safe country” that they reach.Footnote ¹⁰ Governments reinforce safe-country requirements with visa restrictions and penalties on carriers to block direct travel, so that migrants must travel through other countries and seek protection there. The Trump administration issued a regulation to bar asylum applications from any migrants who traveled through other countries where they could have applied for asylum but did not.Footnote ¹¹

Other responses truncate the process for hearing asylum claims or limit where and how asylum seekers may apply. In the United States, asylum seekers at the border or ports of entry are generally subject to “expedited removal.” This means they must, typically without a lawyer’s help, navigate an interview to establish that they have a “credible fear of persecution” before they get a full asylum hearing.Footnote ¹²

The Trump administration tried to deter asylum seekers by detaining them, bringing criminal charges for unlawful entry, and separating them from their children.Footnote ¹³ In 2018, the Trump administration tried unsuccessfully to require that asylum seekers apply only at ports of entry.Footnote ¹⁴ Starting in December 2018, the Trump “Remain in Mexico” policy forced many asylum seekers to wait in Mexico for long periods before their cases are heard in the United States.Footnote ¹⁵ In June 2022, the US Supreme Court found that the Biden administration has discretion to end the Remain in Mexico program, but as of late 2022 the challenge to the program was still pending in a federal district court.Footnote ¹⁶ The Trump administration also invoked a public health statute, Title 42, to cut off access to the asylum system for many forced migrants arriving at the southern border of the United States. In November 2022, a federal district court struck down that barrier as adopted improperly.Footnote ¹⁷

Other efforts to limit asylum have narrowed the refugee definition. Big questions in asylum law are what counts as persecution, how much risk of persecution is required, when persecution is “on account of” protected grounds, and who is ineligible.Footnote ¹⁸ During the Trump administration, Attorney General Sessions issued a decision in 2018 that made it much harder to win asylum based on domestic violence or gang violence.Footnote ¹⁹ The Biden administration vacated this ruling,Footnote ²⁰ but stricter requirements for asylum remain a tool for refugee skeptics.

Interdiction, remote borders, safe-country provisions, and limiting the refugee definition offer destination country governments ways to cast migrants’ claims as beyond legal protections tied to physical presence on national territory. This makes it hard to mount challenges based on civil rights or any other legal framework that relies on some connection to the destination country.

These techniques to manage refugee protection are closely tied to a political strategy that links migrants to a soft altruism that – according to skeptics – must give way to “nation first” to ward off imagined foreign threats. The results are deeply troubling as governments avoid their obligations and act beyond the constraints that the international refugee protection scheme places on unilateral government action. The familiar result is indifferent or cruel treatment of desperate migrants.

4 Refugee Protection as Immigration Policy

Some observers criticize government limits on asylum as blunt instruments to shirk legal and humanitarian obligations (Chimni, Reference Chimni1998: 351). Others have defended measures to preserve the scarce political resource of asylum for the migrants in greatest need (Martin, Reference Martin, Martin and Adelman1991: 30). But the problem is a protection regime ill-suited to migration realities shaped by unsettled political conditions, civil wars, environment degradation, and other causes of forced migration. International and domestic law apply a binary scheme that distinguishes refugees from all other migrants, with life-and-death consequences. To be sure, some cases may be easy to put on the “refugee” end of the spectrum, and other cases more naturally fit the other end for people whose decision to migrate is not forced at all. But in the middle, the line between refugees and other migrants is exceedingly hard to draw, even if the line-drawing process were insulated from political pressure – which it is not. Forced migrants who do not qualify as refugees are still in dire straits, and government responses are typically ad hoc and politically precarious.

Under political pressure, refugee law falls back into the orbit of national immigration policies and politics. Other vehicles for protection that are usually understood as part of immigration law fill some of the gap. For example, Temporary Protected Status (TPS) complements asylum in the United States by protecting many migrants outside the refugee definition.Footnote ²¹ TPS allows some noncitizens to stay if their countries are beset by war, disaster, or similar conditions (Keyes, Reference Keyes2017: 102–107). In 2022, the US government added Ethiopia, Myanmar, Syria, Venezuela, Cameroon, Afghanistan, Ukraine, South Sudan, and Sudan, increasing to fifteen the number of countries designated for TPS.

Debate over TPS reproduces core aspects of US immigration debates. Skeptics object to a de facto expansion of asylum (Jordan, Reference Jordan2018). Supporters emphasize TPS holders’ community ties and contributions, echoing arguments for legalizing the undocumented (Keyes, Reference Keyes2017: 107–112; Motomura, Reference Motomura2014: 181–200). Because TPS fills part of the gap between refugee law and immigration law, the Trump administration’s efforts to end TPS for several countries prompted migrants to apply for asylum and other paths to lawful status (Wilson, Reference Wilson2022).

In this way, TPS is like several other aspects of US immigration law that can help near-refugees – such as T visas for survivors of trafficking, U visas for victims of crimes, and Special Immigrant Juvenile Status (SIJS).Footnote ²² These legal vehicles allow noncitizens to acquire lawful status, and later lawful permanent residence, on humanitarian grounds. A related development is the emerging practice by the US government in 2022 to allow Ukrainians and then Venezuelans to apply in their home countries for permission to enter the United States – though not with formal admission, but instead through its “parole” authority. At the same time, the US government cut off access to asylum for Venezuelans arriving on the US–Mexico border.Footnote ²³ The Central American Minors program is more limited in scope but similarly allows that decisions in home countries may lead to permission to enter the United States (Greenberg, Reference Greenberg2021).

More generally, refugee law and immigration law are never far from each other. Many legal rules combine to fill some of the gap between them. Debates over forced migration follow patterns familiar from immigration law. National immigration policies and politics also pervade debates over asylum. Arguments for defining “refugee” more broadly draw persuasive power from US immigration law’s perceived failures. For example, some asylum seekers have close relatives who are already in the United States without lawful status. If Congress had enacted any of the past decade’s legalization proposals,Footnote ²⁴ many asylum seekers could have joined their relatives via legalization. Though these legalization proposals failed, efforts to pass them fuel political pressure, grounded in overall immigration politics, to define refugee more broadly in the meantime.

Similarly, the perception that forced migrants fall into the gap between refugee law and immigration law can influence outcomes in close asylum cases. Applicants who meet threshold eligibility requirements must also convince an immigration judge or another US government official to grant asylum in the exercise of discretion.Footnote ²⁵ Decisions may ultimately turn on subtle measures of worthiness.Footnote ²⁶ Immigration judges may grant asylum to sympathetic applicants who might fit other avenues to lawful status – except that those routes require long waits. Skeptics criticize such grants as stretching the law and want to limit refugee admissions from outside the country (Rhodan, Reference Rhodan2018). Rebuttals invoke a robust, welcoming nation-of-immigrants self-image. These blurred lines are consistent with the transformation of refugees into immigrants in the public imagination (Alperin & Batalova, Reference Alperin and Batalova2018). In short, debates over immigration policy shape refugee protection.

5 Toward the Fair Treatment of Forced Migrants

Treating forced migrants fairly first requires careful thought about the gap between refugee law and immigration law. The gap is significant. TPS generally does not apply to migrants who arrive in the United States after dire conditions emerge in their home countries. If TPS is unavailable to forced migrants, other available paths to lawful status typically will not assess their degree of peril if they return to their original homes (Keyes, Reference Keyes2017: 107–112). Similarly, unsuccessful asylum seekers may try other paths to lawful status that disregard the harm at the core of their asylum claims.

Perhaps an approach based on national belonging – such as a civil rights framework – should guide protection for forced migrants who do not qualify as refugees. But a civil rights framework is awkwardly suited to assess the claims of new arrivals. Skeptics of immigration may cite this poor fit to argue that governments need not be constrained by the rules that operate inside national borders. But it would be a grave mistake to think that any country can dismiss people who reach its borders but fall outside admission categories and the refugee definition. Migrants will arrive and put political and cultural pressure on any country to find responses that are consistent with its foundational values.

These values do not compel taking in all newcomers. But it is essential to take the needs of forced migrants seriously. Some values will be grounded in humanitarian impulses that drive outrage at the consequences of hardline responses – children separated from parents, or toddlers drowned and washed up on Mediterranean shores or the banks of the Rio Grande. This reaction gains traction from basic ideas: that nation-states share some responsibility for human beings who are displaced and suffering, and that the total disregard of people based solely on their place of birth is an unacceptable affront to human dignity.

From this perspective, human rights can inform protection for forced migrants who do not qualify as refugees. Human rights law, only nascent when the Refugee Convention was adopted, has matured into a broad net of protections. Though human rights are still not directly enforceable in many countries, especially in the Global North, they can play a pivotal role as general principles that can help ascertain when forced migrants who do not qualify as refugees should still receive protection grounded in dignity (McAdam, Reference McAdam2007: 197).Footnote ²⁷

Taking the needs and human rights of forced migrants seriously is all the more urgent if migration is attributable to prior destination country involvement in forced migrants’ home countries. Apart from whether such history creates obligations to accept forced migrants (Achiume, Reference Achiume2019: 1517) there will be substantial domestic and international pressure to respond thoughtfully. Essential is that the overall treatment of migrants should reflect an intelligible rationale and serious efforts to apply legal rules fairly. Though ad hoc, discretionary approaches may sometimes be the best available options, sound decision-making cannot assume this is true. The risk is too great that decisions that are ad hoc or driven by political or opportunism can lead to cruel treatment and mask illegitimate discrimination.

A next step requires recognizing that responses to migration are badly hobbled if they rely on a sanguine belief that the line between refugees and other migrants is objective and immune to political sway, and then make that line hugely consequential (Benhabib, Reference Benhabib2004: 137; Hamlin, Reference Hamlin2021).Footnote ²⁸ Instead, it is essential to see a broad spectrum of migrants with many gray areas and categories that are hybrid and fluid. Relatedly, it is crucial to see forced migrants not just as survivors in flight, but as multidimensional people who will shape the societies where they and their children and grandchildren settle.Footnote ²⁹ Though these migrants have strong needs for protection, they also have a fuller role that includes many contributions, economic and otherwise (Long, Reference Long2015: 13–15).Footnote ³⁰ Just as it is a mistake to draw a sharp line between refugees and other migrants, it is also essential to examine comprehensively each migrant’s future in a society.

Germany, for example, offers forced migrants from Syria more than language instruction and other traditional integration programs. In partnership with the private sector, the German government has tried to draw migrants into apprenticeships that need new recruits, operating as employment-based immigrant admission categories would (Hockenos, Reference Hockenos2018; Siems, Reference Siems2015). Germany has also given unsuccessful asylum seekers from the Balkans special consideration for employment-based categories (Bither & Ziebarth, Reference Bither and Ziebarth2018: 10). Similarly, conditions in countries of origin should influence the design of legalization programs as well as decisions in individual cases involving cancellation of removal, SIJS, or relief for crime survivors. In general, decisions should not only consider the vulnerability of forced migrants, but also harmonize their treatment with other paths to lawful status.Footnote ³¹ Allowing the plight of forced migrants to influence a full range of decisions in their cases would relieve much of the pressure that currently distorts refugee law.

It is essential to assemble ad hoc vehicles such as TPS, T and U visas, and SIJS into a more integrated framework. Extending beyond separate schemes, this framework would reflect some recognition of migration-related human rights, such as the right to life, the right to security of the person, the right to resources for subsistence, and the right not to be persecuted. A useful model is the growing acceptance in the European Union of “subsidiary” or “complementary” protection for forced migrants who fall outside the “refugee” definition, as well as the newly activated Temporary Protection Directive.Footnote ³²

Another key issue is whether protection decisions should be made in individual cases or based on countries or regions. As long as individual cases are decided separately, resources must be committed to ensure that all decisions are careful, with competent legal counsel to keep decisions accurate in light of the facts and law. But separate inquiries might better yield to overall assessments of groups from troubled regions (Aleinikoff & Zamore, Reference Aleinikoff and Zamore2019; Keyes, Reference Keyes2017: 137–147). A group-based approach would reduce administrative costs and delays, and less will turn on luck and access to skilled advocates (Eagly & Shafer, Reference Eagly and Shafer2015: 47–59). The best approach will vary by situation, but there is much room to improve.

Yet another core question is the nature of protection for forced migrants. Is it acceptable for protection to mean only temporary shelter? Is asylum sometimes not the best form of protection for migrants fleeing extreme poverty or environmental degradation? By offering limited protection that does not lead routinely to permanent residence and citizenship, would more people be protected? When is it wise to accept this trade-off and offer less to more people? When is it wise to offer help closer to their countries of origin? But when is that response just an excuse for the tragic and inexcusable failure to help people in need?

Nothing in what I write here is a call to end the essential protections that are based on the Refugee Convention. My plea is for broad agreement not only that those protections are available to too few forced migrants, but also for immigration law to develop new vehicles for admission that go beyond traditional family and employment categories to include forced migrants who do not qualify as “refugees.”

I close by recognizing the limits of these thoughts on the fair treatment of forced migrants. Fair treatment depends ultimately on what can be done to address why people migrate, and why they return – or do not return – to their countries of origin. The fair treatment of forced migrants is unattainable without serious attention to these fundamental issues, which I address in other work (Motomura, Reference Motomura2020: 499–528).

Introduction

Refuge was introduced as a central feature of the post-World War II human rights regime with a goal, among others, of alleviating the precarity caused by forced displacement. Yet only a few decades after those commitments were made, precarity and temporariness were inserted into the very mechanisms intended to forestall lives of uncertainty for forced migrants. Refugee status is difficult for many to obtain and permanent resettlement is often out of reach. Many refugees wait, move, and navigate often indefinite temporariness in camps, without documentation, or inside the same countries where they have been displaced. Given the opportunity, many forge informal bids at permanence when states deny them formal avenues through which to establish permanent residence. Refugees make these informal bids for an array of the citizenship rights associated with full membership. But rights of place – the opportunity to stay somewhere indefinitely, expect reentry if one travels abroad, and move about freely within the territory – are uniquely fundamental to a person’s very existence. Uncertainty about the place from which one exercises all other rights is so disempowering that protecting people from statelessness has always been a central goal of international cooperation in a human rights regime. Among the many urgent dilemmas implied by our conveners’ work – Shachar’s The Shifting Border (Reference Shachar2020b) and Benhabib’s “The End of the 1951 Refugee Convention” (2020) – are the ways that time and spatiotemporal precarity are used to deny migrant persons a pathway to citizenship outside of their country of origin.

While migrant precarity predates the twentieth century,Footnote ¹ the legal means for circumscribing rights to refuge became more deeply entrenched and more diversely structured in the late twentieth century. This entrenchment was part of the process through which all avenues for permanent migration became more closely guarded.Footnote ² Eligibility for protection and for citizenship have come to depend on when claims are initiated, lodging those claims in the right place at the right time, and continual renewal of short-term visas that grant permission to travel in the first place.

Of particular note is the rising prominence of temporary refuge as a means through which states divert forcibly displaced persons into statuses that offer no permanent resettlement or opportunities to adjust one’s status. Temporary refuge exists in ongoing fashion in the EU and the US. Similar programs also crop up on a case-by-case basis in other refugee-receiving countries. In this chapter I consider the challenges associated with temporary protection, paying special attention to what these programs portend for climate mobility, which is almost certain to dwarf all other causes of forced displacement for the foreseeable future. The discussion focuses on unique features of the US program of Temporary Protected Status (TPS). It situates this and similar programs in a larger context of increasing temporariness for mobile persons, noting an “episodic” approach to forced migration. This episodic approach to refuge is an outcome of the genesis of refuge as a response to political persecution but fits poorly with climate displacement. The chapter analyzes what it means to offer refuge on an episodic basis. It concludes by recommending steps for avoiding excesses of precarity and episodic responses to ongoing climate displacement.

1 Time and Temporariness

In migration, time serves as a boundary setting and boundary enforcing technology. Temporal boundaries in the form of deadlines for departure are attached to all visas required by Global North countries as a prerequisite for entry by noncitizens and people without permanent residency. Individual types of visas – for example, tourist, work, and student visas -demarcate a precise time after which a person’s presence becomes unlawful. For the people who hold those visas, time carves out a border that eventually approaches and crosses over them, rendering them unwelcome outsiders after that visa expires. For tourists, students, short-term workers, and others, this is often an undesirable and costly inconvenience. But for people fleeing danger or forced out of their home countries, temporal borders that threaten their legal residence in safe countries stand between them and the fundamental human rights they need to survive.

Refuge and asylum have a special relationship to time and deadlines because forced mobility is an emergency. Refuge and asylum are defined by urgency both with respect to their causes and the legal standards used to decided who is entitled to formal remedies.Footnote ³ The 1951 Refugee Convention and the theories of human rights in which protections for refugees are embedded explicitly identify precarity among the circumstances generating rights to protection. Displacement and statelessness in the wake of World War II and the Holocaust were the reference points for the human rights doctrines that ultimately declared rights of place – residency and citizenship – to be universal human rights. Those rights should, in theory, also obligate states to provide circumstances that protect and rescue people from precarity. In practice, though, this has been a challenge that Global North states fail to meet more often than they succeed. Instead, the arrival and presence of refugees and asylum seekers fleeing urgency creates a sense of emergency for many receiving states to which they respond with varying degrees of closure. Only around 1 percent of all refugees are resettled in a new country each year (FitzGerald, Reference FitzGerald2019). The rest are indefinitely confined to camps or live informally, in cities and in transit.Footnote ⁴ Their precarity is political because they have nothing approximating the political rights that come with citizenship. It is material because most refugees are denied legal opportunities to earn a decent living. But above all else it is temporal because anything they have may only temporarily be theirs. Their future in the place they settle, with the opportunities they make in that place and everything they manage to build there, can be cut short with no warning. All this, amid a life characterized by indefinite waiting and uncertainty, comes to constitute its own form of abuse.Footnote ⁵

2 Precarity and Refuge

Beginning in the 1970s, an additional layer of precarity was added to the refugee experience in the Global North in the form of temporary protection. Temporary protection is a time-limited opportunity for people who cannot safely return to their home countries to remain in host countries. Temporary protection can be renewable but it is not attached to any eventual adjustment to permanent residency or citizenship. Much like short-term work visas, student visas, and other avenues for people to spend extended time in countries where they do not have citizenship, temporary protection is never a direct conduit to a status with permanence. Refuge and asylum offer legal pathways to citizenship. But temporary protection does not.

In the US, temporary protection was formalized as a part of a sweeping 1990 immigration bill that also reorganized immigration priorities, increased the caps on visas for lawful immigration and temporary work migration, and created the Diversity Lottery. In the language of that bill, TPS “[e]stablishes a program for granting temporary protected status and work authorization to aliens in the United States who are nationals of countries designated by the Attorney General to be subject to armed conflict, natural disaster, or other extraordinary temporary conditions.”Footnote ⁶ The status was to be authorized for twelve- to eighteen-month periods and is renewable indefinitely. TPS recipients are generally permitted to work legally in the US and many go on to acquire common markers of belonging and membership. But, significantly, it comes with no pathway to full citizenship or lawful permanent residence no matter how long someone has held TPS.

TPS carves a temporal boundary around people who have been temporarily offered refuge that is slated to end in twelve to eighteen months. As Shachar’s description of a shifting border predicts, TPS’s temporal boundary is almost always in motion, advanced in increments, though never for more than eighteen months at a time. To analogize in territorial terms, people with TPS live as if they are on a very long bus ride ending at the border where they can expect deportation and refoulement. But, just as their bus approaches the border, that border is moved a modest number of miles away from where they had been told it was situated. Their bus approaches the border, the border advances, and the bus continues moving toward it perpetually, never stopping or allowing them to remain in one place.

In practice, this is a bus ride that many never disembark. Today there are people with TPS in the US whose initial designation was granted decades ago. Salvadoran refugees were the first group covered by TPS, and although the initial designation expired in 1996, a new designation was issued in 2001 that remains in effect today, over two decades later. Hondurans in the US received TPS in 1998 and Nicaraguans in 1999, and each has been extended through the present as well. Currently, nationals from sixteen countries hold TPS, many in mixed-status families with spouses and children who are US citizens. In total, there are currently close to half a million persons in the US with TPS (National Immigration Forum, 2022). This stands in contrast to the fact that permanent refugee protections have been offered to fewer than 100,000 people annually in most years since the Refugee Act of 1980 was enacted. In recent years that number dropped well under 50,000 and in a few years almost no one was permanently resettled in the US.

In the EU, a Temporary Protection Directive was adopted in 2001, as a response to the Bosnian displacement in the 1990s. It has only recently been activated for the first time, offering protection to people fleeing Ukraine through March 2023. The EU is currently home to millions more persons with temporary refuge status, as over 7 million people have fled Ukraine in a very short period of time (Karasapan, Reference Karasapan2022). But, even before Putin invaded, there were displaced populations with temporary protection in EU countries. Denmark famously drew sharp critique for withdrawing temporary protection for Syrians who had sought shelter from conflict in their country of origin (Panayotatos, Reference Panayotatos2021). As with TPS in the US, temporary protection in the EU is also not a stepping-stone to automatic naturalization. Outside of naturalization, any rights an alien is granted are always alienable, often on a predetermined schedule and sometimes with the penalty of forbidding return after departure (voluntarily or via deportation). As with many recipients of TPS in the US, although many conjecture that the Ukrainian displacement truly is temporary, no one can know whether this will be the case, how long “temporary” will mean in this case, and who will have good reason to seek to remain past the point when international agencies declare Ukraine safe again. Particularly as the war in Ukraine drags on longer than optimists hoped, people who rebuild their lives in new places may rightly be reluctant to return to a place from which they were traumatically uprooted and where many cities have been leveled.Footnote ⁷

One highly salient distinction between the EU and US versions of temporary protection is the TPS provision that applies to people who are in the US and unable to return home following natural disasters. TPS has been invoked to protect people displaced by natural disasters following Hurricane Mitch in 1998, which hit Northern Triangle countries particularly hard, and again in 2001 and years that followed owing to earthquakes in El Salvador, Haiti, and Nepal. The EU Temporary Protection Directive does not make reference to natural disasters or use other language that breaks with the postwar consensus that refuge is for people who flee conflict and persecution.Footnote ⁸ New Zealand and Canada have created one-time opportunities for very small numbers of qualified persons displaced by climate change to resettle in their countries.Footnote ⁹ In New Zealand a widely-discussed visa for Tuvalu persons fleeing rising water was proposed but failed to gain adequate support. Canada has also given visas to people fleeing storm fallout. Neither country has an established and ongoing program through which people can apply for relief from climate related displacement. Very recently Australia agreed to accept a limited number of Tuvalu citizens seeking permanent resettlement.

Regrettably, despite the uniqueness of the natural disaster provision in US TPS, there is no evidence that the bill’s sponsors in Congress foresaw the need for climate migrant protection or sought to create a precedent asserting any US responsibility for resettling climate change refugees. Indeed, the provision only applies as a form of defensive asylum to people who are already in the US. TPS cannot be sought affirmatively at the border. But, increasingly, climate migration in response to slow and fast-moving emergencies, as well as crises that are forecast but have not yet arrived, is occurring. For this reason, the opportunities and pitfalls created by the US TPS program warrant scrutiny. In the next section I briefly explore the background conditions that led to the creation of TPS, before turning to a normative interrogation of addressing climate displacement with TPS.

3 The Path to Temporary Protection

If one were looking for it, there were signs early in the history of refugee protection that time-constraints were being imposed in ways that made refuge harder to access. The background conditions that led to the creation of temporary protection lie in the 1967 implementation Protocol that followed the 1951 Convention in which the right to permanent refuge was first codified. The 1967 protocol vastly expanded the number and origin points of people who were and would be eligible to claim refuge beyond the very constrained 1951 terms. By the late 1970s, temporary refuge programs began to emerge as a reaction to this scaling up of refugee eligibility (Durieux, Reference Durieux, Costello, Foster and McAdam2021: 679). Initially these temporary programs were described as attempts to slow the pace at which people seeking refuge could claim rights. The stated justification for such programs was to help states manage very large-scale displacements during which it was claimed that it would be impossible to adjudicate claims at the rate they were made (Durieux, Reference Durieux, Costello, Foster and McAdam2021: 679). Australia initiated its first temporary refuge program in 1979 after mass displacement in the Indochinese peninsula (Durieux, Reference Durieux, Costello, Foster and McAdam2021: 679). Beginning in the 1980s in the US and in 2001 in the EU (earlier in some countries such as Germany), informal regimes of temporary protection began to crop up. They were not yet fully codified, but short-term protection was being offered in place of permanent refuge.

This potential for states to use temporal boundaries to throttle and deterritorialize refugee protection had been evident at the very moment that nations committed to refugee protection principles, even as the justification for those agreements was to end precarity for refugee persons. Early refugee protections were transparently circumscribed by dates that referred to displacements of specific groups within specific time windows. The 1951 Convention specifically singles out only causes of displacement that occur prior to 1951 as a qualifying factor in refugee designation (Benhabib, Reference Benhabib2020: 83). Nations such as Germany and Austria that offered the right of return to displaced persons inserted similar temporally circumscribed qualifiers into their provisions.Footnote ¹⁰ These temporal boundaries ensured that only a subset of persons displaced by World War II qualified for refugee status with rights. The claims of other groups facing persecution were delegitimized and the ongoing nature of forced displacement was not confronted. In 1967 the original 1951 temporal markers were overridden. But the larger practice of temporally bounding protections has persisted, and precarity for refugees and other forced migrants remains a likely outcome of displacement.

Temporary protection is not an anomaly in the larger landscape of twentieth-century migration. Contemporary refuge developed in a context of a portfolio of increasingly formalized short-term migration statues. Programs requiring short term work visas, once exceptional, proliferated in the middle of the twentieth century. At times and in select countries they have outpaced programs permitting immigrants to settle permanently.Footnote ¹¹ As Michael Doyle notes elsewhere in this volume, temporary migrants enjoy a few privileges not available to other migrant persons but are also underprotected by international migration norms that were designed to govern permanent migration.Footnote ¹² This trajectory began to emerge before World War II in the US. It became apparent in the 1940s when the H2 visas for Dominican sugar plantation workers and their more notorious cousin, the Bracero program, were initiated. During the postwar period, short-term grants of residency rights for refuge, education, seasonal and skilled work, medical treatment, and other purposes proliferated in the US. For Europe, the decades following World War II were also a time of transformation, during which many countries that a few decades earlier had been countries of emigration were becoming destinations for guest workers recruited by programs in countries such as France and Germany, among others. At the same time that human rights norms encoded in the Geneva Conventions were being put into practice, temporary settlement was also becoming common.

Accompanying the creation of temporary visas and entry in the US during the middle of the twentieth century is a normative transformation in how countries think about temporary residency (Cohen, Reference Cohen2015). Prior to the late twentieth century, temporariness had long been regarded with suspicion as an avenue through which untrustworthy transients might exploit the openness of societies that permitted impermanent or circular migration. But after World War II temporary admissions became the preference of destination countries. The value of conserving and growing the population that once dictated a preference for permanent settlement was replaced with fears about overpopulation, many of which were thinly veiled defenses of white supremacy and resource depletion.Footnote ¹³

During this transformation, established norms prioritizing permanent settlement were replaced with a claim that temporary work migration is mutually beneficial for would-be migrants, host countries, and sending countries alike. It came to be seen as preferable to mass permanent migration. While once intent to settle and naturalize had been an explicit priority for immigration policies, permanence gradually became the prize most fiercely guarded by Global North states. In turn, a temporal form of deterritorialization was established using visa deadlines to perform bordering and exclusion.

Following the transformation of views on short term migration and the accompanying development of short-term opportunities for entry without an avenue for settlement, people with no access to visa-free travel privileges came to rely on short-term visas to exercise mobility rights. Over time, many short-term immigrants were forced to string together multiple status extensions or adjustments of statuses to avoid deportation. These extensions serve as informal versions of long-term residency where permanent residency leading to naturalization and citizenship is unavailable. In some cases, states have been quite openly complicit in this practice. Postwar European countries tried and failed to remove people whom they had recruited for temporary work programs. The US has long had in place several programs explicitly dedicated to promoting indefinite temporariness for undocumented immigrants, for example Deferred Enforced Departure, other forms of parole, and Deferred Action for Childhood Arrivals. Frequently, termination of status and subsequent deportation looms as a very real possibility for any given individual experiencing protracted temporariness. European countries and the US have invested heavily in removing unauthorized residents and in threatening others with deportation as a means of policing borders.

4 Norms and Practices of Temporary Refuge: The Discretionary Episodic Approach

Even in an immigration landscape in which permanent settlement is often foreclosed, TPS is exceptionally difficult to justify. TPS enacts an approach to displacement that emphasizes discretion and treats displacement as episodic. All refuge is to some degree discretionary insofar as states can assess ongoing crises and displacements, designate the refugees they will accept, and revise their caps annually. But a program of refugee resettlement that is adjusted annually (or even an annual allotment of temporary work visas) will be more predictable and less arbitrary than the exercise of discretion on a moment-by-moment basis. Episodic discretion treats the need for refuge as tied to specific events that cannot be predicted and for which remedies can only be implemented post-hoc. Most refugee receiving countries have ongoing visa allotments that commit the country ahead of time to resettling political refugees and do not tie those visas, post-hoc, to any specific event. Episodically defined problems can and perhaps should be addressed with discretionary solutions because single episodes are by definition not generalized.

The circumscription around TPS is more starkly episodic than traditional refuge and asylum because not only does each TPS designation depend on the acknowledgement that a given episode creates an obligation to offer protection to victims, but each episode is also assumed at the outset to be one that will resolve itself. The assumption of future resolution exempts the program from nonrefoulement commitments that make any forced return of people with refuge and asylum unlawful. But it is clear in many instances that assumptions of future resolution are misguided. In the US, TPS is rarely terminated. Even during Trump’s attempt to end TPS and deport people who have been living with TPS for decades, when almost all lawful immigration and eventually short-term travel to the US was halted, ultimately grants for people with TPS were extended and mass refoulement was averted (Rodriguez, Reference Rodriguez2021).

Should climate migration be treated as episodic and requiring episodic redress? Natural disasters, according to the language of the original bill authorizing TPS, are episodes. Like political conflict, storms and earthquakes can be seen as episodic because no two iterations are predictable in their timing or their features. But climate change itself is not a natural disaster. Scientists treat climate change as a phenomenon for which predictive models can and should be consulted.Footnote ¹⁴ Climate scientists are virtually unanimous about the certainty that climate change will continue for the foreseeable future, along with manifestations such as rising sea levels, shifting weather patterns, and changes to food production. Their projections suggest exponential growth in the factors – regional food insecurity, rising sea levels, heat, fire, and so on – that cause climate mobility (The World Bank, 2021). In 2022 the world saw temperatures so high in densely populated parts of South Asia (44–45C/111–113F) that people at risk because of their physical health and/or exposure were subjected to unsustainable, likely fatal, conditions (Ellis-Petersen & Baloch, Reference Ellis-Petersen and Baloch2022). Michael Doyle’s contribution to this volume describes in detail acknowledgment on the part of the United Nations High Commissioner for Refugees (UNHCR) that weather-related displacements are increasing exponentially.

An episodic approach to climate change and climate mobility misrepresents a phenomenon for which predictions already exist. Even though the science of predicting climate change is evolving, scientists agree that climate change is happening, will continue to happen, and cannot be reversed for the foreseeable future (Oreskes, Reference Oreskes2022).Footnote ¹⁵ Displacement is inevitable. A predictable phenomenon for which no remedy currently exists does not call for episodic responses.

Furthermore, climate displacement and migration may sometimes be temporary, but climate science predictions warn that many effects of climate change will render the locations people flee permanently uninhabitable in ways that wars and persecution do not (Lustgarten & Waldron, Reference Lustgarten and Kohut2020). This process has started for people in low-lying islands that are becoming submerged (e.g., Tuvalu) but is likely to accelerate with rising sea levels and temperatures in many regions. At present the largest proportion of this migration has yielded internal displacement.Footnote ¹⁶ But it is difficult to see how these kinds of displacements can remain contained within nation-state boundaries when the scale of inhabited territories being affected by climate change and the speed of that change are outpacing even many dire predictions that scientists offered in the 2010s.

Finally, climate change is also not something humans can control in the way they do wars and cease-fires, even though climate change is a process triggered by human choices. Neither preventive nor mitigation measures can properly be analogized to cease-fires and effective peacekeeping, even if mitigation or preventive measures eventually slow or prevent some displacement. If anything, climate change is less episodic, more predictable, and far less subject to human-made remedy than any form of persecution.

Because climate change is neither episodic, nor unpredictable, and cannot be controlled (in the short term) the way cease-fires and treaties can control conflict, the kinds of judgments that must be made when assessing whether people qualify for temporary refuge do not apply to climate refuge. This is true even in circumstances where persons displaced by discrete events and natural disasters express a preference to eventually return home. Not only do humans not control how the natural processes that we trigger unfold, but also, over time, climate change will remain a poor fit with a discretionary episodic framework for relief.

Furthermore, accepting temporary protection as a default and appropriate response to climate mobilities validates state power to deny important rights to members of their societies, deny them permanent protection, and evade responsibilities for refugee resettlement. Drawing on the concept Shachar advances in The Shifting Border, state responsibility has not been tied to the narrow temporal boundaries of temporary refuge. On the contrary, short-term visas assign their bearers responsibility for their own temporariness. They compel consent to short-term status in circumstances where viable alternatives do not exist. They insist that the displaced commit to leaving when their visas expire, barring a renewal or change of status. But, unlike short term work, travel, and study visas, forcibly displaced persons by definition cannot consent to their displacement. This makes the nature of that consent categorically quite different from the consent of someone whose intentions are to work, enroll in a school, or even visit family.Footnote ¹⁷

5 Pushing Back

In their recent scholarship, Benhabib and Shachar perform the invaluable work of diagnosing important problems with contemporary bordering practices and, crucially, steering scholars toward strategies for moving forward. Shachar emphasizes ways to extend states’ legal obligations so that they follow the shifting border while Benhabib recommends a number of tools including rethinking responsibility and ‘thinking ahead’ rather than treating individual refugee crises as unpredictable surprises with unforeseeable causes. In the spirit of solidarity with their insistence that purely normative work critiquing existing migration protocols will not save lives or bolster the protection and inclusion of forced migrants, I would like to probe how to transform a temporal injustice into temporal justice using widely subscribed principles. This would fulfill Shachar’s and Benhabib’s goals of rethinking responsibility and, in particular, ensuring that responsibility follow the shifting border. Here I want to talk about how to tie state responsibility to time.

The first move forward, following Benhabib’s call to “think ahead,” would require states to move away from exclusively episodic approaches to climate refuge and toward the kind of ongoing resettlement efforts that have been enacted to resettle some portion of people fleeing persecution and war. In some cases this may mean managed retreat but in others it will require more traditional resettlement (Ajibade, Sullivan & Haeffner, Reference Ajibade, Sullivan and Haeffner2020). To be sure, fewer visas for resettlement are made available each year than are sought. But climate migration will not wait for states to fully sort out responsibility for political refugees. Simply acting on our knowledge that climate displacement is certain in the years to come will mark a step toward thinking ahead. While there is good reason for the existing focus on international agreements (e.g., the 1984 Cartagena Declaration), ultimately state definitions of eligibility for resettlement must acknowledge climate displacement.

A second step forward would be to revalue the time of persons who have held short-term visas for long periods of time. This follows Shachar’s recommendation to attach state responsibility to shifting borders. Just as people arriving with permission to resettle would under most circumstances wait a probationary period of time before naturalizing, so too can people with short-term visas eventually have time-in-residence credited toward eventual citizenship eligibility.Footnote ¹⁸ Ongoing adjustment would not be “amnesty,” which is itself episodic relief for people who have lived long-term without authorization papers. Step-up programs toward citizenship acknowledge the undemocratic nature of compelling people whose claims have been validated to indefinitely renew their political status in brief segments of only a few months or years. The explicit justification for the temporary refuge programs currently in place in the EU and US at the time that they were enacted is that some crises quickly displace so many people that even processing their refuge and asylum claims would be impossible for the states where they might make such claims. The plausibility of this justification wanes with each passing renewal.

There are a few concerns that a skeptical reader might have when confronted with an argument to push back against episodic approaches to refuge. First, the fear may arise that in pressing states to adjust the statuses of long-term bearers of short-term visas, states will respond by simply ceasing to offer temporary protection. In other words, temporary protection is the best protection one might reasonably hope states to enact, and the reasons states have for offering temporary protection suggest that, when pressed, they will rescind even this very thin form of refuge. While it is possible that pressing states to do more on behalf of those who they have previously agreed to shelter temporarily could result in reactionary policies, the risk seems low in a context in which states are already generally free to eschew offering any kind of refuge to anyone. This isn’t exactly an optimistic assessment. But Global North countries are increasingly poorly served by mass deportation. Enfranchisement might not be politically popular, but deportation would be much more costly.

Second, it might seem simpler to reject mobility in favor of sedentist assumptions that all short-term mobility is suspect. The related ideals that temporary migration should be exceptional and only migration-as-settlement is acceptable may be well intentioned.Footnote ¹⁹ But short-term migration is now well established. Attaching only weak sets of rights to temporary statuses under the guise of discouraging people from seeking permanent temporariness blames the least responsible agents for temporary migration regimes: displaced persons. Temporary migration is widely known to be the only means by which people, especially people from states in the Global South can find refuge in North America and the EU. States offering temporary statuses are responsible for the terms of those statuses.

Conclusion

The precarity that distinguishes temporary protection stands in sharp contrast to the aspirational twentieth-century ideal that citizenship fulfills a human right to have rights. Over time, the intractability of ongoing and predictable displacement throws into relief the inadequacy of offering temporary extensions of protection to people who are indefinitely, if not permanently, displaced. When and for how long people can access the things to which they are entitled matters. Episodic and temporary protection weaponizes time as an evasion of responsibility for the people most adversely affected by climate change. While short-term visa statuses are now a permanent features of international migration regulations, it they should be treated with skepticism in the realm of refuge and climate migration.

International cooperation for the adequate protection of refugees remains an essential endeavor. Yet its foundations are fragile and, increasingly, its structure is confused. One source of fragility and confusion alike is the web of international agreements that apply the “safe third country” concept, which permits as lawful the return of refugees to jurisdictions where protection already has or might already have been found. Focusing on effects of safe third country rules in European and North American law, this chapter develops a critique of this practice. It does so, in part, on a different plane of analysis than the one that has predominated the literature thus far. While most scholars have criticized the safe third country concept as undermining individual rights protection, I argue that it is implicated in a preceding and more foundational harm: It deforms the possibility of democratic responsibility. I argue that we would do well to see the violations of refugee rights in question as more than privatized harms inflicted on an individual. They are relational and structural wrongs that concern the objective relationships guaranteed by domestic constitutional and administrative law. Perceiving this harm illuminates not only how the safe third country concept has corrupted international refugee law, but also why international human rights should be understood, more broadly, to protect the political agency of democratic citizens. This conclusion might seem counterintuitive in relation to the long-held notion that asserting human rights occurs at a level more abstract and different than, if not opposed to, national schemes of rights protection. But it yields an important analytic shift, in which we see commitments to international human rights and humanitarian ideals to align, constructively and in new form, with the public integrity of democratic states.

*

The safe third country concept has a distinctive genealogy, originating as the “country of first asylum” in Scandinavian legal systems in the 1980s (see Kjaerum, Reference Kjaerum1992; Shachar, Reference Shachar2022). Its principal aim, now more widely codified in bilateral and multilateral agreements around the world, is to constrain the irregular movement of refugees and asylum seekers who have secured or could have secured protection in one country but choose thereafter to travel without authorization to seek protected status in another (see Executive Committee of the High Commissioner’s Programme, 1975–2004b; Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 668). This aim to constrain movement is effectuated in a number of ways at various points and forms of contact between state and refugee: at the admissibility or merits phase of the asylum process as grounds for exclusion; or as legal justification for the interdiction and removal of those attempting entry for the purpose of seeking asylum. The concept is accordingly more than an anchor of asylum policy but informs and enables the border management systems writ large in Europe and the Global North (Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 665). Its consequences for refugee rights, for the posture of the state to refugee claims, and for the coercive measures deployed by the state are substantial and far-reaching.

The view that an asylum seeker may be returned to another state, as long as that state is considered safe, has become all but intuitive, but it is not self-evident. It is worth pausing to reflect upon its presuppositions. To penalize secondary movement is to assert that it is somehow less deserving and more spurious than “direct” movement (Hailbronner, Reference Hailbronner1993). This opens the door to the flawed inference that refugees who move on from the first place of possible safety thereby lose their status and become mere migrants, searching not for international protection but (merely) a better life or economic opportunity; they engage in abusive “asylum shopping” and do not flee as a result of “genuine need” (see Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 669–670).

Accordingly, the safe third country concept permits a state to exclude or remove these individuals without considering the content of their claims to protection. The state assumes a hastened and confrontational posture to the suffering of others, as attention shifts from assessment of refugee status to determining whether the individual can be removed from a jurisdiction safely (Costello, Reference Costello2005). Here, procedures for appellate review are rushed or highly abbreviated, and legal remedies often lack assurance for the automatic suspension of removal decisions (see, e.g., Council Resolution of 20 June 1995 on Minimum Guarantees for Asylum Procedures, 1996, cited in Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 671). Because an individual removed from one state must be accepted into another, the safe third country concept needs to be systematized into bilateral and multilateral readmission agreements that coordinate the forced removal and render it permissible.

Consolidated in the 1990s, the EU’s Dublin Regulation is one such agreement to secure, in its own terms, “effective access to the procedures for determining refugee status” (Dublin II Regulation, Recital 4). It elaborates “the criteria and mechanisms for determining the Member State responsible for examining an application for asylum lodged in one of the Member States” (Article 1) and projects as given that “Member States, all respecting the principle of non-refoulement, are considered as safe countries for third-country nationals” (Dublin II Regulation, Recital 2; Dublin III Regulation, Recital 3). Broadly imitating this architecture, the United States and Canada similarly codified the safe third country concept in their 2002 bilateral Safe Third Country Agreement (Agreement Between the Government of Canada and the Government of the United States of America for Cooperation in the Examination of Refugee Status Claims from Nationals of Third Countries, 2002).

As a contemporary reminder of the concept’s relevance and creative adaptation, in June 2021, the Danish Parliament passed legislation allowing it to establish asylum centers overseas where claims would be processed and asylum itself would potentially be given extraterritorially (BBC News, 2021). Earlier in May, Denmark signed a Memorandum of Understanding concerning cooperation on migration and asylum with Rwanda (The Local [DK], 2021), the first of several possible partner states along migration routes that might host camps and agencies to administer asylum access far from European shores. The United Kingdom in April 2022 proposed, to much controversy and constitutional challenge, to do the same; and only recently has the newly elected Labour government promised to end the pursuit of such plans.

International institutions and scholars have scrutinized the integrity and legality of the concept since its inception (Executive Committee of the High Commissioner’s Programme, 1975–2004a, 1975–2004b; United Nations High Commissioner for Refugees [UNHCR], 1995, 2003), yet the conclusions have been mixed and their emphasis somewhat obliquely placed. Violeta Moreno-Lax (Reference Moreno-Lax, Goodwin-Gil and Weckel2015) makes the revelatory argument that the focus of both scholarly and institutional inquiry has been foremost on the propriety of the concept’s application, with its legality being all but presumed as a general and theoretical matter. The dilemma posed by the safe third country attribution is accordingly understood to concern the particular expectations of what it means for a given country to be considered “safe” in light of obligations under the 1951 Refugee Convention (Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 666; Gil-Bazo, Reference Gil-Bazo2015b: 44). Once this frame of analysis is in place, the only remaining scholarly inquiry is whether such conditions are indeed empirically satisfied – and, if they are not, what legal argumentation can be most effectively crafted to challenge the legality of transfer (see Abell, Reference Abell1999; Mathew, Reference Mathew2003). Some scholars have extended this argument from practicality: Because a removing state has no possibility to guarantee that refugee rights will be protected by a third state, the notion of a safe third country is simply not viable (Durieux, Reference Durieux2009; Selm, Reference Selm2001; see Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 666).

Yet arguments from workability are not yet judgments about essential rectitude or lawfulness. Analyses that critically assess the legality of the concept from a holistic perspective – spanning refugee, human rights, and public international law – are relatively recent and still exceptional (see Gil-Bazo, Reference Gil-Bazo2015b; Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015). Most studies remain equivocal, seeing the possibility of a lawful safe third country concept to be at least theoretically and conceptually coherent (Byrne & Shacknove, Reference Byrne and Shacknove1996; Goodwin-Gill & McAdam, Reference McAdam2007; Kjaergaard, Reference Kjaergaard1994; Vedsted-Hansen, Reference Vedsted-Hansen, Nicholson and Twomey2000). Conclusions tend ultimately toward permissiveness: “[T]he 1951 Convention neither expressly authorizes nor prohibits reliance on protection elsewhere policies” and the safe third country concept, although often abused, is not essentially unlawful (Foster, Reference Foster2007) (see also Hathaway, Reference Hathaway2005, cited in Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015).

These prevailing trends of the scholarly literature have largely traced and informed juridical developments. The United Nations High Commissioner for Refugees (UNHCR) for its part has accepted that returns under the safe third country concept are lawful insofar as the third state affords “effective protection” for the individual returned, which is seen to include nonrefoulement protection, adequate procedural protections for fair and efficient status determination, access to means of subsistence sufficient to maintain an adequate standard of living, and general observance of fundamental human rights in accordance with international law (UNHCR, 1997; see Moreno-Lax, Reference Moreno-Lax, Goodwin-Gil and Weckel2015: 671).

This approach has been endorsed by regional and national judiciaries as well. The European Court of Human Rights (ECtHR) ruled in 2011 that Belgium’s transfer of an Afghan asylum seeker to Greece violated the ECHR’s protections against inhumane and degrading treatment (MSS v Belgium and Greece, January 21, 2011). In light of their nonrefoulement obligations under Article 3 of the Convention, EU member states applying the Dublin Regulation “must make sure that the intermediary country’s asylum procedure affords sufficient guarantees to avoid an asylum seeker being removed, directly or indirectly, to his country of origin without any evaluation of the risks he faces” (MSS v Belgium and Greece, para. 342).

The Court of Justice of the European Union (CJEU) that same year similarly affirmed in the joined cases NS and ME that a member state’s discretion to refuse a transfer under the Dublin system in fact becomes an obligation in cases where rights of the applicant under Article 4 of the EU Charter of Fundamental Rights are in question and when “they cannot be unaware that systemic deficiencies in the asylum procedure” in the receiving state (NS and ME, 2011, paras. 98, 106). Like the ECtHR, the CJEU emphasized that “the presumption underlying the Dublin mechanism, … that asylum seekers will be treated in a way which complies with fundamental rights, must be regarded as rebuttable,” notwithstanding the fact that the “Common European Asylum System is based on the full and inclusive application of the Geneva Convention and the guarantee that nobody will be sent back to a place where they again risk being persecuted” (NS and ME, 2011, paras. 104, 75). Unlike the ECtHR (Hirsi Jamaa and Others v. Italy, 2012), however, the CJEU ruled that “minor infringements” of European asylum law fail the threshold for requiring the suspension of a Dublin transfer and has effectively maintained as its floor the violation of the prohibition against torture (NS and ME, 2011, paras. 82 et seq.).

Early challenges in Canada to its negotiated safe third country agreement with the United States ultimately came to naught before the Canadian Supreme Court in 2009, and present litigation has met only preliminary success in demanding more careful Canadian scrutiny of the adequacy of U.S. asylum procedures for protecting particularly vulnerable persons (see The Canadian Council for Refugees et al v Minister for Immigration and Minister for Public Safety). However, in March 2011, the Inter-American Commission on Human Rights (IACHR), by light of the same reasoning submitted in European jurisprudence, held that Canada’s return of three individuals to the United States under its “direct-back policy” violated the right to seek asylum, protections against indirect refoulement, and rights of due process (Inter-American Commission on Human Rights [IACHR], 2011: 128). Like European courts, the IACHR’s decision turned on the “automatic application” of removal under the safe third country concept and required justiciable assessments of the risks posed to an individual by the other state’s asylum system, of the likelihood of exposure to refoulement, and of the availability of the right to an effective remedy (Gil-Bazo, Reference Gil-Bazo2015b: 73).

How are we to read these lines of jurisprudential development? In one respect, the rulings in Europe and the Inter-American system secured or articulated a higher level of rights protection for those seeking asylum in the procedures of safe third country systems. Following the MSS ruling, for example, most EU member states indeed suspended their transfers to Greece. After the decision in NS/ME, that halt became nearly universal, and the prohibition on transfers to states with “systemically deficient” asylum protections was later incorporated into the Dublin Regulation’s recast, Dublin III (Arts. 26, 27 and 29) (see Fratzke, Reference Fratzke2015).

Yet, in another respect, the responsibility for a state’s coercive measures and the integrity of human rights protection continues to be in essential matters vague and contested. While Dublin III reflected the CJEU’s language in NS and ME, it failed to provide further administrative or legal guidance on the meaning of “systemic deficiencies” that require prohibition of transfer. Substantially divergent interpretations among member states of what “systemic deficiencies” entail have meant that states have hesitated to apply the prohibition to other cases that nevertheless bear similarity to Greece – including inadequate processing and reception capacity in Italy or Bulgaria (Fratzke, Reference Fratzke2015). Dublin’s recast erected a new legal presumption that forestalled critique of state practice and instead served to entrench state prerogatives anew (Linden-Retek, Reference Linden-Retek2021).

This retention of state prerogative soon carried over to the EU itself as a state actor. In 2016, the EU released a statement in which it formalized with Turkey a readmission agreement to return any “irregular migrant” found to have entered the EU through Turkey without already having applied for asylum. Also eligible for return are those who submit claims to asylum in the EU but have arrived from any other safe third country or first country of asylum where they could have previously received protection (European Council, 2016). As part of the agreement, the EU committed in exchange to increase resettlement (one for one) of Syrian refugees residing in Turkey; to accelerate Turkey’s negotiations to accede to the EU and the visa liberalization process; and, finally, to offer additional financial aid packages to support Turkey’s aid to Syrian refugee communities.

In 2019, the Trump administration responded to the substantial rise of Central American asylum applications by seeking bilateral “asylum cooperation agreements” with Mexico, Guatemala, El Salvador, and Honduras (O’Toole). At the same time, circumventing conventional statutory requirements for negotiating the full terms of safe third country agreements, the Trump administration adopted new asylum regulations unilaterally. On July 16, 2019, the US Department of Homeland Security posted an interim final rule (IFR) on Asylum Eligibility and Procedural Modifications that prohibited asylum applications at the southern border from those who transited a third country on their way to the United States but failed to seek asylum there (Asylum Eligibility and Procedural Modifications, 2019). In February 2023, the Biden administration announced plans to renegotiate a “transit ban” with Mexico and to reinstate bars to the right to seek asylum for those refugees who pass through yet fail to apply for protection in third countries (Miroff, Sacchetti, & Sieff, Reference Miroff, Sacchetti and Sieff2023).

These developments suggest that safe third country arrangements are remarkably resilient and, further, that analyzing their harm in the frame of individual rights protection risks a rather limited form of critique. The concept’s complicated resilience suggests, specifically, that a renewed verification of individual protection risks formalism; it addresses only part of what makes the safe third country framework harmful and only part of the role that principles of human rights and refugee law arguably should play in a normative order of global justice and accountability. A focus on the effect of the safe third country concept on individual rights is an analysis of what would make the concept lawful in application, but this does not yet speak to the lawfulness of the concept’s philosophical and political foundations.

The Danish example is in this regard telling, for it is not immediately apparent as a matter of substantive protection why such measures would necessarily violate human rights commitments. One can indeed imagine, given the wealth and expertise of the Danish state, that sophisticated, well-funded measures might be established such that the minimal human rights of refugees could indeed be satisfied elsewhere. In principle, the prospect is not out of the question. And yet the threat of such an arrangement to the spirit, purpose, and structure of the international human rights regime would, I suspect, remain acute in the minds of many advocates and scholars (see Motomura, Chapter 1; Schmalz, Chapter 4; Mégret, Chapter 5). But why, exactly? What is the source of this intuition? Answering this question, I argue, requires a new theoretical perspective. The individual rights frame of analysis is, on its own, insufficient.

Focusing on the safety of the transfer, as the individual rights frame would have us do, concedes too much to the worldview that posits the state as a stable point of origin – for analysis, normativity, and decision-making. It accepts categorically that the state is in the first instance entitled to decide to initiate a transfer, in light of its own determinations of the safety of others and what this safety requires. While these determinations might be frustrated or proven incorrect should safety not in fact be guaranteed, the presumptive structure of that decision – the sovereign prerogative to impose a vision of safety upon others beneath the veneer of international asylum law – is not similarly frustrated. The rectitude of the state’s desire to transfer is left unchallenged by this law; and states have continued to act upon this desire, irrespective of and unconcerned with its rectitude.

The harm of this pretense is that it permits a deeper form of withdrawal from the space of responsibility for human welfare and human recognition. Appreciating this harm requires a distinctive philosophical and analytic frame, one I call the frame of democratic responsibility.

The anxiety one feels when contemplating safe third country schemes stems only partially from the human rights violations they foreseeably might cause. It stems also from an antecedent, more foundational concern: the notion that an exercise of discretionary coercion might instrumentalize life for the purposes of the state and thereby render the agency of that life illegible to the state. This concern is more foundational not because it necessarily contemplates greater physical consequence to the asylum seeker but because, before its physical consequences are evident, it distorts and prejudges the relational perceptions at work in the encounter between the state and the refugee (see Moreno-Lax & Lemberg-Pedersen, Reference Moreno-Lax and Lemberg-Pedersen2019). As effects of this relational harm, the individual rights violations that might come do not reveal the most vital dynamics of the concept – nor why, even should its applications be made to satisfy individual rights minima, it would remain immoral and unlawful.

This argument rests not on a theory of domination or legitimate coercion but, instead, upon a theory of democratic agency and democratic personhood that, in taking the task and burdens of such agency seriously, binds the fates of the citizen and refugee together. The deception that ensnares the refugee is in truth a self-deception that ensnares the state. The dilemma posed is in one sense far narrower than consideration of the adequacy of “protection elsewhere”; it is oriented closely to the relationship between the state and the refugee. But in another sense it is thereby far broader – leading to consideration of the structures of power and relations of influence that characterize and inflect the encounter between state and refugee. In centering the immediate moral dilemma posed to any bounded political community by the claim of another’s personhood, this analytic frame better apprehends the moral complexity that attends the “porousness” of borders – and it suggests why refugee law is a body of law directed not simply at “others” – to them “outside” – but also to the polity, to “us” already here (see Linden-Retek, Reference Linden-Retek2021).

Witness to immensely turbulent periods of forced human migration, Hannah Arendt argued in 1951 – precisely at the birth of the Refugee Convention – that the international system and the modern democracies that comprise it fail to secure for large numbers of people of the earth the right to belong to a political community, one so organized such that they are treated as equals within it. This right was so foundational to democracy and to humanity that Arendt famously termed it the “right to have rights” ([1951] 2004: 376). The new political institutions that were consequently created to manage statelessness and migration – from the 1951 Convention through to Dublin and the American asylum cooperation agreements – speak to this fact, and to its endurance. Arendt’s words suggest that it is far from coincidental that modern polities struggle with the integrity and meaning of their borders or continue to find the refugee and the migrant making appeals to safety and inclusion. The refugee and the stateless person are not victims of natural scarcity but products of a particular political form: the state whose people imagine themselves to be sovereign, separate from others, and sovereign strictly by virtue of their separation (see Näsström, Reference Näsström2014: 547). The problem that results, inevitably, is that when one loses, for whatever reason, membership in one’s own country, there is nothing to guarantee that one will secure admission to another. “What is unprecedented,” Arendt writes, “is not the loss of a home but the impossibility of finding a new one” ([1951] 2004: 372).

Political theorist Sofia Näsström has argued, in light of this problem, that we ought to see the “animating principle” behind the “right to have rights” as the burden of responsibility. “The reason,” she writes, “is that it is only by sharing this burden that human beings can take it on, and this is precisely what membership in a democracy does. By making us into an equal among others it limits and defines a responsibility we cannot plausibly shoulder on our own” (Näsström, Reference Näsström2014: 547). To be deprived of the right to have rights – to be stateless – is to be deprived of this common experience of assuming human responsibilities; that is, of the political structures that make the burden of responsibility possible.

Because the human capacity for the assumption of responsibility can take hold only among one’s fellows, those deprived of political community will be “perceived as the most irresponsible person on earth” (Näsström, Reference Näsström2014: 560, emphasis in original). The stateless are “the absolutely innocent ones,” Arendt writes, “and it is precisely this absolute innocence that condemns them to a position outside, as it were, of mankind as a whole” (2003: 150). Unable to apportion responsibility for one’s acts among others intelligibly, one is in effect removed from the realm of humanity altogether; one is dehumanized: To be stateless, of course, is indeed to be human, but it is to exist without the means to live out one’s humanity.

Seeing the connection between humanity and the shared burdens of responsibility helps explain Arendt’s valorization of membership in a political community as something essential not just to democratic self-government. It is through citizenship that we remain human beings who articulate and make sense of our lives through action and judgment in concert with others (see Näsström, Reference Näsström2014: 560–561). Arendt writes that the refugee has indeed been deprived of shelter and of security, of the protection of the state; but foremost the refugee’s deprivation is of “a place in the world which makes opinions significant and actions effective” ([1951] 2004: 372–376). It is a fascinating formulation because it suggests that the loss of human rights signals a deeper void within political life that implicates more than the individual whose particular rights are at the moment in question. It is a problem in which others are implicated, as well. For it is they who decide to acknowledge another’s opinions and actions – or not; which is to say, they decide whether they, too, will live out their humanity, or not.

Following Arendt, Étienne Balibar (Reference Balibar2014) and Seyla Benhabib (Reference Benhabib2018) have elaborated the fragile, iterative imbrication of freedom and equality, captured in Balibar’s term égaliberté and Benhabib’s ‘democratic iterations’, by which the inclusion of the stranger denotes the very possibility of creating institutions which would establish and recognize equality. This possibility restores to political agency its attunement to the democratic burden of responsibility. The “right to have rights” is a reminder of – and a means to affirm – this relational form of agency (Benhabib, Reference Benhabib2018: 103–109; Gündoğdu, Reference Gündoğdu2015). While implicating the state’s legitimate claim to democratic authority, it reaches beyond the state as an image of peoplehood and beyond statist institutions that claim the exercise of public power. The right might be addressed formally to the state’s agents, its border guards, asylum officers, or immigration judges; but it attaches, too, to the broader civil engagements of citizens: those who act solidaristically without prior state sanction, perhaps, or in generative networks of mutual aid and legal assistance. Such acts call to mind, contest, and accordingly expand the consciousness of the democratic society for which the state otherwise seeks to speak, on whose behalf it purports to take responsibility.

A key consequence follows: Absent the moment when the state can respond to the refugee and her claim to asylum, both are in fact in a position of irresponsibility. The refugee, owing to the deprivation of a political community; and the democratic state, because denying human beings inclusion in political life degrades the “normative basis of democracy” (Näsström, Reference Näsström2014: 561). The state, too, risks presuming to place itself outside human power and law – and thus losing its own humanity. Conversely, when the state attempts to sustain, revive, and respond to the agency of the other, it restores that humanity, democracy’s normative basis. What refugees seek is the assumption of human responsibility. Counterintuitively, this is what they can themselves provide to those from whom they seek it. This is the philosophical frame of democratic responsibility; in it we find the nature of relational harm and the possibility of relational repair.

Reflecting on Arendt’s work, Itamar Mann in his study of maritime migration and the foundations of international human rights, writes that speech is the “most rudimentary expression of freedom”; “[refugees] must be able to speak who they are” (2016: 130). In the possibilities of speech there lies that spontaneous, unpredictable transition from the bare life of stateless irresponsibility to the birth of potential membership in a community. At stake in the refugee’s agency is therefore a relational responsibility of those who hear the refugee’s claim. Speech as freedom requires that someone listen. It requires a relationship to be established. Freedom is, with Näsström, a democratic quality. While it exceeds membership in any one particular state, it remains predicated on a democratic form of community – and thus on finding democratic relationality anew. This new community must therefore understand itself to be responsible for the agency of the refugee, not merely for her protection (see Mann, Reference Mann2016: 130). It requires sensitivity to the conditions under which her estrangement from responsibility might fall away.

The implication is that protecting individuals from refoulement – the heart of the individual rights concern with the safe third country concept – concerns not only their survival but also their agency. What safe third country arrangements risk at the most fundamental level is this relational harm – a disregard for this dimension of responsibility and what it requires.

With this concern in mind, we can better trace – and more fruitfully critique – the mechanisms by which the safe third country concept deforms the democratic freedom of refugees. If the reception of refugees is perhaps the paradigmatic encounter (see generally Mann, Reference Mann2016) when the burden of democratic responsibility is disclosed, the application of the safe third country concept compromises this disclosure. So how, more precisely, does this deformation of the character of human agency occur?

In the first instance, we can scrutinize the subtle but consequential shifts in asylum procedures that alter the nature of the speech that is heard. Rights to due process under asylum procedures – those ensuring translation and transcription, the right to request additional time and to appeal, the right to request the presence of a doctor or social worker, the right to legal assistance, and so on – align well with protecting the agency for those seeking asylum. Indeed, the structure of the interview, normatively reconstructed, aims to preserve and protect the narrative intelligibility of the story the asylum seeker tells. It structures a narrative space. The encounter is not merely with a human being in pain, but a person with a story and the capacities to tell it, to hear our own, to intertwine their tale with ours.

This space of agency contracts, however, along with its procedural rights, when the safe third country concept is employed. The questions one asks when the safe third country concept operates are simply not those one would ask without it being in place. In different ways, these distortions enforce crucial disjunctions in the refugee’s story – and thus undermine the integrity of her agency. There are two variants of this, characteristic of different implementations of the safe third country logic. The first is the inclusion of a prior admissibility interview; the second is the raising of the standard of proof for successful asylum claims. Let me detail each variant and its consequences in turn.

Illustrative of the first variant are the terms and purposes of the readmission agreement between the EU and Turkey. Under the EU–Turkey deal, the safe third country concept requires a prior admissibility interview before what would otherwise be the eligibility interview inquiring into the grounds for asylum (see European Parliament & European Council, 2013). The question of admissibility tracks whether the state receiving the application claims is indeed the one lawfully responsible. It thus becomes a precondition for any further review of the asylum claim. What is asked of asylum seekers in an admissibility interview, however, changes the posture of the state to their claims and privileges certain lines of inquiry over others. The procedures that previously were meant to secure the asylum seeker’s narrative agency now displace such agency.

Admissibility questions asked by EU asylum officers first concern the conditions of the applicant in Turkey and why the applicant felt it necessary to leave Turkey for the EU. The disposition of the interview – understood to scrutinize the appropriateness of applications in Europe – is to rebut the presumption that Turkey is a safe third country. It is not on the merits of the claim to asylum. The European state here assumes a defensive posture, not an exploratory one; and the burden of proof is placed on the applicant to establish the inadequacy of conditions in Turkey and whether there is indeed reason to fear persecution, serious harm, or the risk of refoulement there. At no point is the asylum seeker asked about the conditions in her country of origin and the many reasons she initially sought the protection of the international community elsewhere. The focus is to scrutinize her “secondary movement.”

Secondly, admissibility interviews also focus on an applicant’s general vulnerability (Zimmermann, Reference Zimmermann2015), namely whether they fall into certain categories that would make them eligible for special protection notwithstanding their movement from Turkey. This question is important, but it also comes with its own risks, being posed as it is apart from the broader narrative of the asylum seeker’s decision to flee. It can easily be a dangerous exercise in dehumanization and, in some cases, of retraumatization of the applicant. For some categories of protection that can be ascribed with less difficulty (unaccompanied minors, handicapped or elderly persons, pregnant woman or those who recently gave birth, for example), this concern is perhaps less grave. But for those seeking protection under categories concerning victims of torture, rape, or other kinds of assault or exploitation, including human slavery or trafficking, the process of proving oneself worthy of this protection by affirming one’s severe vulnerability is immensely fraught.

The point is not that one should avoid speaking of vulnerabilities; these are indispensable to understanding the import of admission. But there is a danger of perceiving asylum seekers merely as victims when such admissibility questions are taken apart from the eligibility review through which the applicant could share their full story. As Arendt understood, for the retelling of vulnerabilities not to reproduce the characterization of “bare life” (Agamben, Reference Agamben1998; Arendt, [1951] 2004: 302), they must be held close to the political collapse (the loss of home) that is their ultimate cause. Vulnerabilities spoken about in isolation from the context in which they are felt and lived risk essentializing asylum-seekers’ experiences and their personhood.

The second variant by which asylum procedures under safe third country logics displace refugees’ agency is in the shift of the standard of proof and the more stringent dismissal of claims. Consider here the Trump administration’s modification to asylum procedures in the United States with its 2019 IFR. The IFR claims to conform with obligations under international refugee law because it presents no bar to the right to apply for withholding of removal under the Immigration and Nationality Act (INA) (IFR, 2019: 834–835). The INA’s withholding of removal standard requires an applicant to make a showing that it is “more likely than not that he would be subject to persecution” in his country of origin (Huang v. Holder, 2014: 1152). Under the prevailing interpretations of the INA, however, asylee status requires merely the lesser showing of a well-founded fear of persecution or past persecution, which entails demonstrating “to a reasonable degree that his continued stay in his country of origin has become intolerable to him for the reasons stated in the definition [of a refugee] or would for the same reasons be intolerable if he returned there” (Immigration and Naturalization Service v. Cardoza-Fonseca, 1987: 439). The result is that the IFR’s imposition of a higher standard of proof effectively risks denying asylum and perhaps even protection from refoulement to a refugee otherwise able to satisfy the “well-founded fear of persecution” standard normally required (see United Nations High Commissioner for Refugees [UNHCR], 2019). This elevated threshold is more consequential given that many asylum-seekers filing applications in the United States are not afforded legal representation.Footnote ¹

While the EU–Turkey deal effectively foreclosed initial discussion about the reasons for fleeing the country of origin, the IFR commits the opposite distortion: It heightens the import of the reasons for fleeing, while neglecting the reasons for not seeking asylum in a country of transit. That the latter could and should have happened is presumed beneath the categorical determination that transit countries are safe. Such juridical forms that raise the standard of proof conceive the personhood of refugees in an equally impoverished light, often in terms of what Didier Fassin calls “humanitarian reason,” which scrutinizes precarious lives, emphasizing and evaluating the likelihood of their suffering (Fassin, Reference Fassin2012; see also Gündoğdu, Reference Gündoğdu2015: 78ff, 111, 157). The higher this bar is, the more exaggerated its effect in reducing the political subject to the status of victimhood alone. This obscures the needed acknowledgments rooted in history and politics – those historical entanglements and particular responsibilities of public institutions. It obscures what the state owes and why it owes it.

The foregoing analysis has attempted to reorient the debate about the legality of the safe third country practice in international asylum law. It has parsed the legality and morality of the concept by formulating a critique within a new frame of analysis: democratic responsibility. The relational parsing of harm suggests far more serious deformations to human personhood than were legible to the analysis of individual human rights violations. Moreover, as failures to take up the burden of responsibility, these deformations implicate the personhood of both citizen and refugee.

My argument here is inspired by the exhortation of Paul Weis that “[t]he development of the law on asylum is inextricably bound up with the general development towards the greater recognition and protection of the human rights and fundamental freedoms of the individual by international law” (1966: 194, cited in Gil-Bazo, Reference Gil-Bazo2015a). In offering a critique of the manner by which democratic states fail to affirm the worth of human agency, I have aimed to say something about international human rights as a political and democratic project in the register of responsibility in a world of shifting borders.Footnote ²

More than a project of claims to individual rights, human rights seek at heart to correct the failure of the modern democratic polity to honor its own vision of responsible political life. By seeking to control and truncate the agency of the refugee, we also inhibit our own. This is why we should understand human rights to have democratic ends. The normative horizon of international human rights law is not, as counterintuitive as it might be, the protection of the individual. Instead, it is the protection of the democratic judgment that we make as members of a nascent, ever-anticipated community of humanity. While the protection of human life is necessary for this judgment, it is not its exhaustive purpose. For human life finds meaning in the human capacity for action and understanding – the ability to take responsibility, among others, for what one does. This renders the individual no less valuable; quite the opposite, it finds within the individual the expression of all political possibility: hope for ending the conflict a refugee flees, for perceiving the many dangers that attend her journey, for making a new country a new home; even for judging what it would mean, finally, to make a third country safe.

The different treatment of Ukrainian refugees in Europe, in comparison with other groups of refugees, has been the object of much debate recently: Is the higher willingness to welcome refugees from Ukraine in comparison to those from African or Middle Eastern states due to the geographic proximity? Does it relate to the circumstances of the war that affects Europe more immediately than other military conflicts?

There are legal factors to the different treatment, especially the fact that Ukrainian citizens already enjoyed visa-free travel in the Schengen states for up to ninety days, which meant that crossing the border and entering the European Union was possible for all those in possession of a passport. Yet beyond the legal situation, the political reaction clearly differed. European Union member states quickly decided to activate, for the first time ever, the Temporary Protection Directive.Footnote ¹ And the public support for refugee reception, at least in the first months, was exceptionally high, with widespread civil society initiatives.

Witnessing the reception of Ukrainian refugees, some commentators have stressed the positive example that should inform refugee protection more broadly, while others have expressed criticism about the inequality of treatment. What is clear is that instances of discrimination (Akinwotu & Strzyżyńska, Reference Akinwotu and Strzyżyńska2022) at the borders are inacceptable. More intricate is the question if the different treatment of refugee groups is justified. Some have argued that it reflects the role of racism in the perception of the war and in the attitude toward refugees (Połońska-Kimunguyi, Reference Połońska-Kimunguyi2022; Ramasubramanyam, Reference Ramasubramanyam2022). Others have called the reference to Ukraine being in the direct neighborhood of the European Union a “fallacy of geographic proximity” (Lacy & van Houtum, Reference Lacy and van Houtum2022).

This is but one recent example of discussions about the legitimate response of states to different groups of forced migrants and, more generally, about conditions for a legitimate regime of international protection. A core aspect of these debates is the role of geographic proximity. Is the proximity of refugees’ places of origin a legitimate criterion for different access to, and levels of, protection? The role of geographic proximity is a familiar object of debate in refugee law also from another angle: The basic rule of nonrefoulement is interpreted by several states to only apply if the refugee arrives directly from the state they flee from. Together with visa policies and carrier sanctions, this has limited protection obligations largely to neighboring states of refugees’ places of origin. This restrictive interpretation of nonrefoulement through the safe third country principle is criticized as harming the overall function of the international protection regime.Footnote ² It leads to problems of responsibility-sharing among states and contributes to increasingly hostile conditions for those seeking protection.

Here, too, the legitimate role of proximity is at stake: Is proximity a workable criterion for distributing responsibility among states? Can such restrictive interpretation of nonrefoulement be reconciled with the overall goal of an effective protection system? Whereas the debate about Ukrainian versus other refugees concerns the legitimacy of more favorable treatment, the debate around the scope of nonrefoulement concerns the legitimacy of states seeking to limit their obligations.

Against the background of these debates, the present contribution seeks to systematically explore the role of proximity and arguments about its legitimacy. The first part analyses how physical proximity of a state of origin and of a person’s location matter in international refugee law and in human rights law regarding refugees. It observes how states have increasingly narrowed down obligations in interpreting refugee law, and how de facto the ability to “come into proximity” of a host state is often decisive for persons’ access to protection. In a second part, the chapter then asks about the conditions for claiming and enforcing rights. Rights guarantees are object of constant processes of interpretation, especially so rights at the border. The scope of state obligations toward forced migrants is continuously contested, and even where rights are recognized, it is difficult to hold states accountable for violations. The physical location of claimants affects those conditions of making rights effective.

Overall, the first two parts show that proximity is significant for migrants’ rights to international protection and underline how this role of proximity bears problematic effects on the system. A third part turns to considerations about the legitimacy of the proximity criterion. Proximity can be, at its basis, a reflection of the concrete link that assigns universalist obligations to particular states. This requirement of a concrete link will be hard to forego in a world structured by territorial political communities and limited freedom of movement. Yet proximity is by far not the only possible way of establishing a concrete connection. For a viable system of protection, states must base their responsibility on these different forms of concrete links. The role of proximity does not have to be negated entirely, as long as it is coupled with effective access to territory and protection, and complemented with other bases for state obligations, taking into account links of causation and refugees’ explicit choices.

1 How Proximity Impacts Protection: International Refugee Law and Human Rights Law

International refugee law and human rights law operate as exceptional constraints to states’ otherwise discretionary rules on admitting noncitizens. We can call these rules the law of international protection, which includes foremost the 1951 Geneva Refugee Convention with the 1967 Protocol, and human rights guarantees that are applicable in the context of forced migration. The law of international protection is concerned with rights of persons vis-à-vis states they are not a member of. In that sense, it structurally raises questions about how universalist rights can be translated to obligations of particular states. How to assign or distribute the responsibility for protection is an issue not explicitly regulated: Refugee law and human rights law contain rules about the rights of persons and state obligations toward them, but not about the relationship between states in securing those rights. This distribution of responsibility results indirectly through the application and interpretation of the laws of international protection.

2 Claiming and Enforcing Rights: The Quandary of Distant Claimants

Proximity plays a role not only at the foundation of what is considered a right to protection but also in the conditions for claiming rights. This is true, first, on the level of claiming rights through legal procedures. Access to procedures is not formally dependent on the location of claimants; in practice, however, it often is. Persons whose rights were potentially violated will have difficulties to challenge acts in court if by the very same acts they were physically expelled or returned to another place (Spijkerboer, Reference Spijkerboer2018: 464). It is telling that nearly all significant cases before the ECtHR regarding migrants’ rights in recent years resulted from strategic litigation.Footnote ¹⁹ In other words, it was not migrants themselves who decided to litigate against state agents who had rejected them at the border without procedure, or expelled them from the territory without identification.

Apart from factors such as lacking resources to approach lawyers and to find one’s way through national and international court procedures, there are more specific aspects that complicate the access to court procedures for migrants at distance: A system that requires individual, clearly identifiable claimants to assess the legality of state acts also means a dependence on documentation. In many instances, while the state acts are of continuous and documented nature, it is a major challenge to prove that a specific person was subjected to that treatment at a specific occasion. In the procedure before the ECtHR, this requirement forms part of the condition of victim status. Proving their presence at a specific time and incident is often difficult for those who by the acts in question were removed to another place.

Notably, distance also affects the processes of claiming rights politically.Footnote ²⁰ In scholarship on the foundation of rights, often centered around the Arendtian notion of the “right to have rights,” this role of proximity has remained largely unaddressed. In recent years, there have been important contributions underlining the role of political struggles for human rights, especially in the context of migration. These contributions sought to explain how universal rights remained precarious despite the existence of international human rights treaties. They shifted the focus from a discussion on the foundations of human rights to the practices of political founding (Butler & Spivak, Reference Butler and Spivak2007: 44; Gündoğdu, Reference Gündoğdu2015: 171). Historically, it took political movements and experiences of prior rights violations for the adoption of rights treaties, and for persons gaining recognition as rights holders. Also once achieved, rights require political support to remain effective.

In the context of migration, this focus on “politics on the streets” is particularly relevant given the structural lack of citizenship and therefore a lack of access to institutionalized forms of participation (Johnson, Reference Johnson2014; McNevin, Reference McNevin2011; Ahlhaus, Chapter 15). The political claiming of rights is in that sense linked with questions of visibility and collective action rather than formal vote. However, those political struggles depend, in general, on presence. The idea that rights are founded and claimed through political action tends to rely on a picture of persons already being copresent.

Jacques Rancière has offered an influential conception of the process of becoming a political subject and a subject with rights. According to Rancière, the possibility of politics is grounded in the capacity of speech and appearance. While the persons excluded from political action may first not be recognized as “capable of speech” (Rancière, Reference Rancière1999: 22), through the acts of “claiming rights they have and do not have” (Rancière, Reference Rancière2004: 305), persons constitute themselves as political subjects. His conception relies on a picture of politics as “climbing the scene” (Rancière, Reference Rancière1999: 25). While being of explanatory value, this also raises the question how or whether such processes of gaining rights can take place in constellations of absence and distance.

Migrants claiming rights from afar lack physical presence and to some extent also visibility. Those kept at distance through acts the legality and legitimacy of which stand in question, can rarely address the responsible state organs, and the responsible publics, directly. There are mediated forms visibility, through public media or decentralized ways of communication, that can mitigate the effects of distance. Yet the absence of personal encounter impacts the possibility to claim rights. Martina Tazzioli (Reference Tazzioli2015) has criticized the focus on appearance in public as a “citizen-model of politics.” The model illustrates that copresence, or proximity, is often at the basis of politics (cf. Waldron, Reference Waldron2011). Is the “presence-model of politics” just one that begins from the concrete interpersonal encounter rather than a given definition of what counts as political questions? It is hard to deny, however, that border regulations are genuinely political questions: They determine the delimitation of, and access to, the political community; they concern those included and those excluded. As such, the quandary of rights of those still distant, or removed to distance, remains.

3 Proximity, Affinity, Vulnerability: Grounding Specific Protection Obligations

It has been shown in Sections 1 and 2 that proximity constitutes a central criterion for assigning protection obligations to states, and that this focus on proximity creates problems of distribution and of effective access to protection. In this section, I argue that, nonetheless, proximity should not be condemned as an entirely illegitimate criterion but rather be complemented with other bases for protection obligations. In developing this argument, I look at different approaches from legal and political theory that discuss the basis of protection obligations. Two qualifications are due: First, it is my premise that states regulate immigration and territorial access. There are notable approaches which dispute that states are free to limit territorial access (e.g., Cassee, Reference Cassee2016), but I will not discuss those and start from the situation of controlled international borders and limited mobility rights. Secondly, my interest is not justifications of protection rights as such, but more specifically the discussion about toward whom a state has obligations. In a world of limited mobility, criteria are necessary to link refugees’ rights to protection with obligations of specific states, and my focus is on what those criteria should be.

At the outset, proximity reflects a basic link between persons, and between a person and a state. In that sense, Jeremy Waldron suggests that proximity is the basis on which people form political communities (Waldron, Reference Waldron2011: 2). His account is based on a reading of Kant, holding that human relations, and consequently law and politics, emerge from people sharing a space. To think of rights and politics as based on physical proximity contrasts with accounts of communities based on shared religion or history. Waldron in that sense juxtaposes his conception with a conception of community based on affinity. In practice, most states will have elements of a shared history that connects people, but also major elements of a community based on living together. The vast majority of states allow noncitizen residents to become naturalized after a certain number of years.

This says nothing so far about protection obligations. In Waldron’s understanding, it does not follow from the principle of proximity that states could not control entry: Even for a community built on proximity, self-determination is relevant, and with that, rules on membership and entry (Waldron, Reference Waldron2011: 18). What can be drawn from the understanding of political community as based on proximity is the idea that proximity is an enabler of political relationships.

Looking at an understanding of community that puts more emphasis on affinity, we can turn to the writings of Michael Walzer. His understanding of political communities being based on shared histories and culture also informs his proposition regarding criteria for protection obligations. Walzer stresses that states have obligations toward refugees, detailing some cases of when such obligations arise. At the center of his account stands the notion of “affinity.” First, affinity results – and the most far-reaching duties – from having caused what turned persons into refugees (Walzer, Reference Walzer1983: 49). In addition to this “affinity by causation,” Walzer sees special duties based on ideological as well as ethnic affinity toward persons in need of protection. Beyond those relationships of affinity, he recognizes a right to protection, but argues that such right cannot be enforced against particular states (Walzer, Reference Walzer1983: 50). Walzer acknowledges the prohibition to return people once they have found safety.

In Walzer’s account, geographic proximity does not explicitly matter for special obligations to offer protection. While he includes the principle of nonrefoulement, he does not primarily envisage a system where refugees seek protection by arriving at the border. Instead, he stresses elements of causation, and of ethnic, religious, or ideological affinity.

David Miller offers an alternative conception which puts the choices of refugees and the notion of vulnerability at the center. Miller starts from a collective responsibility that states have toward refugees, a responsibility which is then assigned to particular states. This happens, he suggests, through the acts of refugees themselves, “making a visa application at a distance, […] turning up at the border, or […] entering illegally and then asking for asylum” (Miller, Reference Miller2016: 83). While this may seem arbitrary since it does not distribute responsibility equally among states, Miller argues, it is not unusual that responsibilities arise in a somewhat arbitrary manner. What establishes in his view the specificity of the claim of refugees (unlike other people whose human rights are at risk), and at the same time the claim against a particular state is their act of “making themselves vulnerable” to that state (Miller, Reference Miller2016: 84).

The concept of vulnerability has an established place in the context of refugee protection, for instance regarding prioritized resettlement or relocation, and with view to positive human rights obligations (for a general account Besson, Reference Besson and Burgorgue-Larsen2014: 75). Specific in David Miller’s conception, however, is that it is not the vulnerability as such but the combination with an act of those seeking protection. Refugees have a claim against the state they have approached “by virtue, first, of having established a physical connection to that state and, second, of having become vulnerable to the decision” that the state will take in response (Miller, Reference Miller2016: 85). In that sense, Miller’s approach differs from one that would view vulnerability as the better alternative over the place of arrival for prioritizing refugee claims (e.g., Welfens & Bekyol, Reference Welfens and Bekyol2021). Miller’s conception relies not on a given proximity but on a created connection, which can be based on arrival at the border as well as on an application over distance.

Itamar Mann’s concept of the encounter goes in a similar direction. Mann’s approach is reconstructive; he does not describe an ideal regime but identifies threads of meaning from historical developments and legal cases. His focus is not on legislative bodies or courts in creating law, but on situations of border controls and enforcement, in which human rights are negotiated. In those moments of encounter, especially at sea, Mann (Reference Mann2016: 48) argues, migrants invoke their humanity, together with an act of putting themselves in the hands of those who can save their lives (Mann, Reference Mann2016: 47). Those situations of encounter move the acts from the level of state responsibility to the interhuman level. Mann’s approach binds back the role of physical proximity to the foundational moment of human encounter. Rather than an abstract criterion why a state should have an obligation toward a particular person, proximity is the reason why a person feels a concrete obligation. In that sense, it shows how proximity is decisive – not original proximity of the state of origin but the physical proximity in a moment of encounter. Unlike in Miller’s conception, the act of making oneself vulnerable here does not focus on vulnerability arising from dependence on state decisions but more physically on the dependence on an immediate act. In that sense, the physical proximity plays a role. While many types of encounters can take place without physical proximity – mediated by videos, calls, writing – the resulting link of dependence differs. In that sense, Mann’s reconstruction underlines the role physical encounter can have in making rights and obligations real and robust.

Mann’s account is closer to the existing legal regime, focusing on rights at the border. Yet it shows the link to the political dimension of claims to protection. While these can be legally assessed, they are also politically negotiated, by bringing refugees’ motives and actions into the focus. With this link to the political, the approach relates to Waldron’s principle of proximity. None of the mentioned approaches argues that the proximity of a person’s state of origin should be the main criterion for offering protection. However, Waldron’s general account of proximity as a mode of founding a political community and Mann’s concept of the encounter show how far proximity and contact can be relevant criteria.

The approaches of Walzer and Miller offer cues for which additional criteria could be viable in grounding protection obligations. Neither vulnerability nor affinity seems to be a workable criterion alone, and the authors do not suggest this. In Miller’s account, a person seeking protection choses the contact with a particular state, and this can happen at the border as well as from afar in an embassy. While courts have rejected the derivation of such a system from existing human rights obligations,Footnote ²¹ state legislators would be free to establish procedures to apply for humanitarian visa at embassies. Certainly, the reception of refugees from distance, whether through special visas or through resettlement, is an important addition in an international protection system that strives for better responsibility-sharing between states (Doyle, Reference Doyle2018a; Schmalz, Reference Schmalz2019). Such a system cannot rely solely on financial contributions of states but must create avenues for refugee reception in other than the neighboring states (Aleinikoff & Owen, Reference Aleinikoff and Martin2022: 464).

Finally, Walzer’s notion of “affinity by causation” can be read as a radical idea for advancing reception obligations that account for historical injustices, be they colonial histories, military operations, or involvement in the destruction of the environment. Tendayi Achiume has coined in that connection the idea of “migration as decolonization,” which includes economic migrants (Achiume, Reference Achiume2019). There are increasing efforts to find legal answers to the growing phenomenon of displacement in the context of climate change (McAdam, Reference McAdam, Costello, Foster and McAdam2021). While it is difficult to single out states in that regard, as at least industrialized states share the responsibility for climate change, the focus on causation can nonetheless inform the debate. In that sense, protection obligations based on affinity should be seen as an important addition, which opens up new forms of mobility and responsibility.

Conclusion

International protection is largely guided by a “rule of proximity”: States seek to limit secondary migration and to keep refugees in the countries bordering their country of origin. Human rights obligations toward migrants are limited to those who make it “into proximity” too. The interpretation of jurisdiction has followed an idea of states’ responsibility to begin with physical, rather than administrative or other indirect, control over persons. Distance affects not only the legal rights to protection, but also the capacity to protest against rights violations: Those who were illegally pushed back or denied access often lack the avenues for claiming their rights. Overall, the current restrictions of protection obligations are deeply problematic. Limiting nonrefoulement to states of refugees’ first arrival might be in line with the wording of the 1951 Refugee Convention but it is not in line with the underlying idea of effective protection and a system built on cooperation between states.Footnote ²² Narrow interpretations of human rights standards, some states’ open disregard for international legal rules, and the limited scope of control have given room to frequent violence and impunity at borders.

The picture of the current regime with the rule of proximity at its core is somber. Yet what are alternative criteria for structuring protection obligations in the current world of limited mobility? The chapter has argued in Section 3 that we should not abolish proximity as a criterion for protection altogether. At the outset, the current system is dysfunctional not because it allows persons to claim asylum when reaching the territory, but because it allows states so generously to hinder persons to reach the territory (Benhabib Reference Benhabib2020; Shachar Reference Shachar2020b). In that sense, the concept of seeking protection at the border should be complemented by additional forms of grounding protection obligations. This could happen along an interpretation of jurisdiction that includes forms of indirect control and thereby extends the human rights obligations of states toward migrants. It could happen through legislative acts that introduce access to humanitarian visa at embassies. And it could add causation as a criterion for protection obligations, viewing forced migration as embedded in a global context, in which states have contributed to reasons for flight.

Introduction: Borders of the Mind

The force of the border, such as it is, is not only that it potentially stops migrants from entering, but that it also builds migrants in its image. The border has long ceased to be a mere physical border. Rather, it is a series of technologies deployed to regulate the movement of peoples across space that deeply structure social and legal relations and are internalized by mobile humans. It moves with them, against them, and sometimes for them (Shachar, Reference Shachar2020b). As such, territory is redefined as a series of pockets of immobility that constrain the outflow and inflow of persons. More importantly, the border shapes what I would describe as a series of mobile identities, the identities one needs to endorse in order to pass the border, conceived in increasingly narrow terms (Sajjad, Reference Sajjad2018). In effect, the border not only regulates bodies, but it also produces particular kinds of subjecthood and subjectivity. These relate, for example, to claims, often reductive and simplistic, about the conditions in the country one left (Kelley-Widmer & McKee, Reference Kelley-Widmer and McKee2024) as well as the particular reasons one feels it necessary to invoke to justify one’s mobility.

The harshest effect of the border is surely on “bare migrants” who cannot claim to be refugees, most notably those “forced migrants” fleeing poverty, ecological degradation, and crime – but not specifically persecution in the sense of the 1951 Convention. If refugees have the right to asylum, then some otherwise deserving migrants will want to qualify as refugees, at some cost to themselves and to the system. Quite aside from populist rants about “fake refugees,” it is inevitable that some migrants will be tempted to claim asylum, especially in a context where what constitutes persecution may be ambiguous and where the narrowness of asylum is all too obvious. This includes migrants pretending to be children when they are not (Gibb, Reference Gibb2018); pretending to be part of a family unit when they are not (Burnett, Reference Burnett2019); pretending to be of a different nationality than they are (e.g., Syrian rather than Moroccan or Turkish) (George, Reference George2015); and criminals or terrorists posing as refugees (Reuters, 2016).

By contrast, those who can claim persecution are a priori and paradoxically in a privileged position (Price, Reference Price2006) in that they at least formally have a right to asylum protected under international law. The relative privilege, for the purposes of cross-border mobility, of refugees has of course become somewhat theoretical as states have deployed considerable resources to make it difficult for even would-be refugees to seek asylum (Gammeltoft-Hansen, Reference Gammeltoft-Hansen, Chetail and Bauloz2014). This has been in part because of a suspicion that many asylum seekers are, in fact, “migrants in disguise,” but also quite simply because of states’ lack of enthusiasm to comply with their 1951 obligations even in relation to bona fide refugees. Still, asylum seekers are historically and legally privileged subjects when it comes to cross-border mobility in a context where states maintain a strict upper hand over who crosses their borders.

Nonetheless, even asylum seekers feel the particular pressures of the border. Whereas the border forces some migrants to claim asylum, this chapter is interested in the converse phenomenon, namely how asylum seekers are under pressure to tone down the extent to which they are also, in some respects, migrants, lest this compromise their bona fides as refugees. In seeking to uphold a category of asylum seekers fleeing persecution as entirely distinct from that of migrants, the international legal regime of cross-border mobility does violence to the more complex reality of mobile humans’ trajectories. Specifically, it confronts them with the choice of one of two stereotypical roles: on the one hand, the bare human fleeing the immediate threat of persecution at the expense of all else for mere protection; on the other hand, the fully agentic but unprivileged migrant.

It thus forces refugees into a bind: either seek protection in the closest or first “safe” territory that they can access by prioritizing their most immediate need for protection, but at the cost of frustrating their life plans; or “migrate” to the place of their choosing to seek asylum but then risk reinforcing the suspicion that they are, indeed, migrants. This tendency is reinforced by an evolution of the conception of refugeehood away from a more political emphasis on human dignity toward a more humanitarian vision, and away from an understanding of asylum as a right toward an understanding of asylum as a favor (Owen, Reference Owen, Oliver, Madura and Ahmed2019) that imposes its own pressure to conform. Even the “privilege” of refugeehood, it turns out, is paid at the price of the voluntary suppression of one’s agency, including as it might otherwise manifest itself as a project of full and long-term resettlement in the asylum state far beyond the immediate needs of protection.

In this chapter, I argue that the iron discipline of the border requires asylum seekers to play the part of the “good” refugee (a refugee defined by pure protection needs) (Szczepanik, Reference Szczepanik2016). This not only creates a constant culture of bad faith, but it also risks the constant erasure of their agency, notably as it expresses itself in strongly entertained preferences to seek asylum in some countries rather than others. The insistent focus on the protection of refugees, however well-intentioned it may be, ends up imprisoning refugees in a unidimensional identity, that of the performance of their humanitarian frailty. It makes incomprehensible why they would take considerable risks to seek asylum in some countries that are removed from the closest “safe” country. It also makes their specific needs in terms of the eventual location of their asylum “unsayable” for activist groups supporting asylum seekers’ rights who may be wary of blurring the distinction with migrants. Eventually, this plays into the hands of governments vying to make asylum seeking difficult, and ordinary migration even more so. In short, refugees are called upon to symbolically reinforce the very border that they seek to traverse – enlisted in their own oppression, as it were, and in breaking solidarity with their fellow “migrant” travelers who cannot rely on a compelling narrative of persecution.

The chapter begins by highlighting why the agency of asylum seekers in relation to the country where they wish to emigrate is impossible to entertain within the categories entertained by the border. This is reinforced by a politics of defending asylum seekers that constantly reinscribes their difference from migrants – and only manages to protect the former at the expense of the latter. The chapter goes on to show how the preferences of asylum seekers cannot be addressed within the law, which at best offers a very limited and contrived defense of their ability to request asylum “in the country where they find themselves.” Contra this invisibility, the next section emphasizes the obvious sociolegal point that asylum seekers are more than just asylum seekers; namely that, woven into their demand for protection from persecution, are multiple aspirations to gain asylum in particular countries that are also so many normative claims against the border and the obligation it imposes to be either a refugee or a migrant. In the conclusion, the chapter speculates about the legacy of a border ideology that severely constrains asylum seekers’ ability to articulate their mobility as a political claim.

1 The Invisibility of Refugees as Migrants

Consider the explosive debate that started in 2022 following the UK government’s project to send asylum seekers who make it to Britain across the Channel to Rwanda. The project has been described by Human Rights Watch as “cruelty itself” (Ahmed & McDonnell, Reference Ahmed and McDonnell2022). But what exactly is wrong with that plan? Many arguments that it will be expensive or inefficient in deterring smuggling are not particularly principled or based on what would be best for refugees. It may be that the British government is seen as skirting its asylum responsibilities, but in a way it is clearly deploying considerable resources to allow another (apparently willing) state to do so. But the prevailing discourse suggests something more complex, namely a genuine reluctance to send asylum seekers to countries other than the one where they have already secured a foothold as a result of their own, often considerable efforts. What is specifically “cruel,” then, about sending asylum seekers to another country than the one they have managed to reach?

That deportation to Rwanda is not per se the problem is shown by the fact that it is more generally third safe state agreements that have been heavily criticized (Linden-Retek, Chapter 3; Schmalz, Chapter 4). For example, activists have long railed against Australia’s arrangements with Nauru and Papua New Guinea to screen asylum seekers on the Manus Islands. Similarly, the EU’s project to ask the government of Niger to establish frontier zones on African soil has been roundly condemned. Again, though, it is not clear from a protection standpoint what is wrong with being resettled in Nauru (as distinct from being detained in Nauru in squalid conditions, evidently), for example, rather than Australia.Footnote ¹ Nauru may not be as wealthy or welcoming for refugees as Australia (although hasn’t Australia amply shown on some level how unwelcoming it in fact is?) but maybe it is “safe enough” for the purposes of bare protection compared with daily bombardments in Aleppo or Taliban repression?

To be clear, it remains possible of course to challenge safe third country discourse on protection grounds alone, by pointing out that some such countries are, in fact, not safe, or at least not quite as safe as they appear to be. Third countries might turn out to be less than ideal from the point of view of asylum itself. It may be that such schemes are part of an ill-thought-out plan to simply “dump” asylum seekers in countries from which they will promptly leave anew and therefore evidence both a dereliction of duty from host states and ultimately unworkable policies. Israel’s infamous practice of sending Eritreans and Somalis to Uganda and Rwanda with $3,500 cash in their pocket but no official status led many to eventually leave again (Gidron, Reference Gidron2018).

But in many cases third countries will indeed be relatively safe, including for asylum seekers. Notably, the reasons for individual persecution in refugees’ state of origin will often no longer be present. For example, although it has been pointed out quite correctly that Rwanda does not have a perfect human rights record (although nor does the UK, to be clear), the gravest violations it has been accused of are against long-time dissidents and activists heavily involved in the contestation of Kagame’s stranglehold on power. It is not particularly plausible that the Somalis, Yemenis, or Iranians who might end up living in Rwanda would be on the receiving end of that kind of dissident-oriented repression. There is, maybe, just a touch of racism involved, as if an African country could not possibly be a safe haven. Indeed, on protection grounds alone, all of these destinations (Rwanda, Nauru, etc.) may be as good as any. Offshore processing and safe third-country agreements in principle at least ensure that asylum claims are examined, as opposed to practices of pushing back refugees so that they may not even seek asylum (Gammeltoft-Hansen, Reference Gammeltoft-Hansen, Chetail and Bauloz2014).

Finally, the issue is not only with poor countries acting as safe third states. Famously, in Calais, hundreds of migrants have crossed the whole of Europe only to then risk their lives daily to cross the channel to the UK. Calais has often created puzzlement. Why would one risk one’s life as an asylum seeker to reach the UK? Surely refugees are no more likely to be persecuted in France than on the other side of the Channel and are, effectively, quite safe by the time they have made it that far. France may be a country with many flaws, but it is a stable and rich democracy with a well-developed rule of law and rights protections. Do not those individuals believe their claim to being asylum seekers by seeking to enter the UK illegally despite already being in a country where they are objectively safe?

This does not mean, to be clear, that it is a good idea to send Somalis, Yeminis, or Iranians to Rwanda – but not necessarily for protection-focused reasons. It is almost as if governments had chosen to take international refugee law at its word by insisting that they are taking protection needs – albeit only protection needs – seriously. Troublingly, then, it is the one-dimensionality of protection discourse itself that paves the way for governments to promote first safe country logics. After all, if basic protection is ensured, then any safe third state is equal to any other. Safe destination countries are interchangeable. It is that element of the border discourse that makes it incomprehensible and even suspicious that Iranians or Kurds, having made it all the way to Germany or France, would risk their lives on a dinghy to cross the channel. And while there is an obligation under the Refugee Convention to process asylum claims on one’s own territory, it is hard to see how this should be incompatible, in principle, with a distribution, including a fairly authoritarian one, of the burden of processing and welcoming refugees (after all, it is not particularly fair that refugees end up predominantly seeking asylum in some countries rather than others, for example through geographic accident).

2 The Politics of Protecting Asylum Seekers

Instead, what these examples suggest is the inability of conventional asylum seeker discourse to provide a strong account of what is wrong in such situations and, in particular, to ascribe any particular normative value to refugee choice of country. The elephant in the room, however, is the suspicion that asylum seekers also express strong preferences to seek asylum in certain countries rather than others (as explored in more detail in Section 3) – that is, that they want to both seek asylum and immigrate. That agency, however, is what cannot be expressed under the conditions set by the border because of the way it might undermine the notion that refugees are a special kind of migrant defined almost exclusively by their flight from persecution. This could then dangerously feed into state discourses all too willing to consider that refugees have all along been, in fact, would-be migrants.

This means that asylum seekers themselves need to tone down their own agency and essentially play the part of the “good” refugee (a refugee defined by their pure protection needs) (Szczepanik, Reference Szczepanik2016). Asylum seekers’ portrayal by sympathetic constituencies, it turns out, increasingly relies on “perceptions of helplessness, suffering and ‘deservingness’” (Neikirk, Reference Neikirk2017: 63) that objectively marginalize refugee agency. This toning down of agency is also evident in the long-standing construction of the asylum seeker as a victim of circumstances, as clearly emphasized in the slogans of the United Nations High Commissioner for Refugees (UNHCR) that “No one choses to become a refugee”, and “People become refugees not by choice but by circumstance” (United Nations High Commissioner for Refugees [UNHCR], 2022c).

Such statements are problematic given the absence of incompatibility between being propelled to move by circumstances and the significance of choice: Refugees, needless to say, are not automatons but living subjects who ponder the very difficult decision to depart and then exercise multiple forms of agency on the way including, as it turns out, about their hoped for final destination. Having minimized their agency at the point of departure, however, it should come as no surprise that the refugee paradigm does not elaborate on that continued agency en route to the final destination. This is consonant, in fact, with a deeper aversion to representing refugees as historical and political agents as a result of a “totalizing narratives of humanitarianism” (Sigona, Reference Sigona, Fiddian-Qasmiyeh, Loescher, Long and Sigona2014: 378).

The irony, then, is that even as some activists may be wary on a deeper principled level of the refugee/migrant distinction, they are often driven to reinstate it on a strategic and political level to comply with the sprawling exigencies of the border. Most are committed to the notion that we should at least protect refugees and that it would be dramatic if asylum seeker protections were watered down even more significantly than they already are. This then makes it difficult to speak about asylum seekers’ agency in a context where anything that might draw attention to their simultaneous “migrancy” could compromise the imagined purity of their asylum seeking or even endanger the “asylum seeking” category as such (“bogus refugees”). Even “pro-asylum” organizations thus tend to reproduce a discourse of the refugee as a universal dehistoricized category of humanity (Malkki, Reference Malkki1995).

In turn, this has two self-fulfilling if unintended consequences. First, it further plays into government efforts to strongly draw a line between refugees and migrants since the two are constantly being pitted against each other, in large part and ironically of course already to comply with the border’s dichotomous language. Second, asylum seekers themselves, having been so typecast, may internalize the border’s exigencies and foreground their protection needs at the expense of any sense of their life strategies. The refugee label becomes “a master status that defines a person above and beyond any other form of identity” (Marlowe, Reference Marlowe2010). This makes their otherwise expressed agency about choice of country difficult to explain since it cannot be invoked in its own terms, even as it transfers the production of knowledge about the refugee to Western humanitarian “experts” (Rajaram, Reference Rajaram2002).

3 Refugee Choice of Asylum Countries under International Law: Neither Prohibited nor Allowed?

This invisibility of asylum seeker agency is, crucially, not just the result of policies but inscribed in the law itself, so that attempts to seek clarity in the international refugee legal regime tend to be inconclusive at best: The law is part of the border, just as the border is part of the law.

It is true that the Refugee Convention protects individuals who have secured a foothold in one state from being deported to another without consideration of their claim. Nor is there anything that suggests that asylum seekers should as a matter of international law claim asylum in the first safe country. That is clearly a late development that is not borne out by the 1951 Convention itself. A refugee may seek asylum in whichever country she happens to be able to do so. As a British court put it in Ex parte Adimi, therefore, “some element of choice is indeed open to refugees as to where they may properly claim asylum,”Footnote ² in at least the very bare sense that the Convention is silent on the issue. An asylum seeker may thus cross a safe country and nonetheless decide not to make an asylum claim there. In a sense, then, by default at least, the Refugee Convention protects asylum seekers’ de facto choice of a particular country to hear their claim.

But nor does the Refugee Convention particularly facilitate the ability to choose the country in which one seeks asylum. It merely seems to conceive of individual as, at best, “popping up” in countries where they then proceed to claim asylum and have a right to have that claim heard in situ. The normative value of such arrangements seems largely pragmatic and functional. Indeed, although deportation tends to have a bad name (not least because of its historical associations), this is not so much because of deportation itself as because of where one is being deported to. Consider, for example, the possibility that the deportation regime between the UK and Rwanda emerged the other way round, namely that asylum seekers found in Rwanda were, following an agreement with the UK, to have their asylum claims processed there. Leave aside for the moment the geopolitical improbability of such an arrangement: Would we still insist that Rwanda process the asylum applications at the expense of the UK because of a fetishization of the place wherein a claim was first made? Would we do so, moreover, even in the face of evidence that asylum seekers were, in fact, delighted that their applications would be processed in the UK rather than Rwanda? In short, our problem seems to be less with deportations of asylum seekers in and of themselves than the fact that deportation may in some cases run against the better wishes of asylum seekers, in ways that are not accounted for by the dominant model.

As to the argument that refugees may demand asylum in the country of their choosing, although it is a proper construction under international law, it may sound normatively quite thin. Refugee law might still be understood to militate against excessive risk-taking in seeking asylum in that there is something contradictory about fleeing great persecution risk only to then further “needlessly” endanger one’s life in seeking to make an asylum claim in one country rather than another. Moreover, although there is clearly preference in Article 31 of the Refugee Convention for not penalizing asylum seekers for entering the country illegally to claim asylum, this is only if they are “coming directly from a territory where their life or freedom was threatened” and provided they “present themselves without delay to the authorities and show good cause for their illegal entry or presence.” This suggests an emphasis on asylum as a form of immediate protection that is not a particularly good description of asylum seekers who have crossed the whole of Europe to, for example, then seek asylum in the UK.

In fact, even if asylum seekers are not violating international law by choosing where they seek asylum, this does not mean that they cannot otherwise be removed to a third safe country. The Refugee Convention, in particular, does not necessarily prohibit administrative practices such as the European Dublin rules, which allow states to send back asylum claimants who have already been fingerprinted in another EU country to that country. Nongovernmental organizations sensing that the “safe third country” conceit is just an excuse for states to opt out of their asylum obligations – but wary of producing any discourse that might hint at refugees being migrants – have sometimes tackled the practice merely on protection grounds. For example, Pro Asyl argues that: “The very idea that a country could be deemed safe is absurd. Simply taking the European Union as an example, who could seriously argue that some minorities do not currently suffer from discrimination, accompanied by sometimes deadly violence, within its borders?” (Pro Asyl, 2016). That argument, however, seems confused. If there are no safe countries at all, then what is the point of the refugee regime? Clearly, many and probably all (European) states have problems of discrimination, which are nonetheless not sufficient to automatically make them unsafe for the purposes of asylum. The argument seems to claim too much and to miss an opportunity to cast the wrong of expulsion for what it is.

In short, the Convention certainly does not prohibit choice of country where one seeks asylum but nor does it facilitate it, except at best through its silences. Refugee law, one senses, has no strong theory of why it might be appropriate to allow asylum seekers to choose the country where they apply, except through the relatively narrow route of stopping states from going on a slippery slope that leads to the avoidance of their refugee Convention obligations. This ultimately leaves where one manages to seek asylum to chance, including geographical proximity and accessibility, financial means and connections, networks, age/gender/race, and so on. It also encourages risk-taking and trafficking networks. Most importantly, it is normatively impervious to the very real agency that asylum seekers express when they seek to obtain asylum in some countries rather than others.

4 And yet They Chose: Refugees as Migrants

The singular focus on protection as the only goal of asylum seekers flies in the face of more complex understandings of refugee motivations and trajectories that simultaneously challenge the border’s hold on imaginations. Refugees are not solely defined by the fact that they seek protection. They are complex agents with complex backgrounds and ambitions (Barsky, Reference Barsky2017), for whom fleeing conditions of persecution, obtaining protection, and securing prospects for a decent life are all intertwined as part of processes of “mixed migration” (Kane & Peterson, Reference Kane and Peterson1995). Refugees may in some cases have wanted to migrate all along, and persecution merely have been a decisive factor in triggering departure; or they may have never thought of emigrating before encountering persecution but been immediately thrust through their flight into a web of complex life decisions about where to ultimately seek asylum. But migrate they do, in a context where refugee flight “represent[s] one variant of the more general phenomenon of migration” and “involves many of the considerations other migrants go through when physically disconnecting from home regions or countries” (Essed, Frerks, & Schrijvers, Reference Essed, Frerks, Schrijvers, Essed, Frerks and Schrijvers2004: 7).

The point seems to be that, like ordinary migration, the decision to seek asylum is driven by push and pull factors and is, ultimately, a real decision, albeit one evidently constrained by circumstances: One would not, for the sake of argument, seek asylum at any cost anymore, by contrast, than one would do so for merely futile reasons. Clearly, if one’s expected eventual situation abroad were worse than the dangers to which one is exposed at home, then seeking asylum would not make sense. Of course, the problem is compounded by the fact that the decision to flee involves a complex form of risk management in conditions where one never knows fully the exact risk of harm at home, the likelihood of being able to seek and obtain asylum abroad, or the kind of life that one may eventually have as a refugee. The element of rationality involved, however, is not nullified by the fact that in practice the choice of destination is heavily constrained by misinformation, opportunity, and criminal exploitation (Gilbert & Koser, Reference Gilbert and Koser2006).

What does characterize refugees is the fact that they have departed a state because of a well-founded fear of persecution and therefore in conditions that leave less time for the deliberate planning of emigration – yet that does not defeat the fundamentally migratory nature of asylum seeking. In all cases and in a context in which flight and asylum seeking do not simply “happen” but typically require complex organization, courage, and drive, the decision about where to ultimately seek asylum is at least as important as the decision to flee and seek asylum at all. Moreover, that decision-making process is rendered even more pointed by the many efforts deployed by states to make it difficult to seek asylum: This means that refugees can leave very little, in fact, to chance. In particular, the effort to enforce a first-safe-country rule creates an incentive for asylum seekers to make sure that they are not “stuck” in a country which is less than ideal for their purposes, and therefore to redouble their efforts to target a particular country for asylum.

In practice, the decision to seek asylum in some countries rather than others is framed by a series of considerations. The primary one is of course seeking a place where one will be free and safe from persecution or harassment, either from the long arm of the state of origin and also at the hands of the host state. Considerations of relative ease of access and likelihood of obtaining asylum will also feature prominently. But factors more characteristic of any migration process will also determine the choice of country. These include prospects for gainful employment and social integration. Clearly, not all safe states are equally attractive when it comes to opportunities (in that respect the UK may well prove more attractive than Rwanda for many asylum seekers), and even generously granted asylum may lead to a life of de facto economic and social misery. The job market is a factor, as is the possibility of having one’s studies recognized, the existence of welfare services, and the availability of low-skilled work for the unqualified. Cultural factors may also have a role, notably language. To flee the danger of persecution to find oneself in a perpetual state of economic and social alienation is not an attractive prospect.

Finally, asylum seekers have tended to be drawn to countries where, quite simply, they already have personal links and where they may be reunited with family, relatives, community, or more broadly a diaspora for sentimental but also practical and economic reasons. A key for understanding refugee choice, then, is that it is a form of inchoate family reunification, albeit operating wholly aside from the legal category of family reunification. More generally, asylum choice reflects patterns of “chain migration” that depend on kinship networks. Although diasporic studies are sometimes dissociated from the study of immigration, in practice the “pull” factor of already established diasporic communities is a strong one. One European study found that the “network effects” involved in some countries already “having become popular destinations for some time” are the single biggest predictor of choice of asylum country (Neumayer, Reference Neumayer2004, 176). In short, refugees seek to both obtain asylum and to make sense of their life trajectories in ways that may only seem difficult to scrutinize as a result of the disciplinary biases of the law-as-border.

5 Arguing for Refugees as Migrants

One would not think less of an asylum claimant’s claim to be a refugee simply because that claimant was not content with seeking asylum in the first safe country, given the obviousness of the ties that bind to some places rather than others. In taking risks to reach certain countries for protection not only on protection grounds, refugees fully express, de facto, their dual identity as asylum seekers and migrants. They also express, evidently, legitimate human aspirations – just not the ones that the international refugee regime is trained to have its sights on. In that respect, asylum seekers arguably do not merely seek the border: They also call it to account as a particular political-legal choice. The claims that are expressed in the process of crossing borders can be understood as not just strategic but also normative.Footnote ³ They are a challenge to the border itself, and its tendency to prevent asylum seekers from choosing a country that fulfills their aspirations as migrants.

To ignore or deny refugee agency is, at the most basic level, to fail to treat refugees seriously as moral agents seeking the best lives for themselves. States that deny that agency add insult to injury in a sense, compounding an initial persecution in the state of origin by denying asylum seekers the ability to remedy it to the best of their abilities in another state. As such, they fail in honoring a deeper cosmopolitan obligation of hospitality that is also a test of their moral integrity. By highlighting that part of refugees’ trajectory is based on a series of voluntary choices, conversely, one allays fears that to give effect to these choices is to perversely do the persecuting state’s bidding, while tapping into a rich register of liberal reasons why one would want to recognize mobile humans’ “life plans.”

Moreover, as David Owen has argued, there may be deeper sui generis grounds for acknowledging “refugee choices […] as having normative significance, namely that such recognition of refugee as autonomous sources of claims to justification is the normative basis of the political institution of refugeehood.” In such a situation, “[t]o exclude or ignore refugee choices is to fail to exhibit the moral respect for persons expressed in the claim to equal political standing (as effective membership of a state) of persons that the institution of refugeehood is designed to uphold in the face of state failures” (Owen, Reference Owen, Oliver, Madura and Ahmed2019: 36–37).

The denial of refugee agency, moreover, is made even worse when it coincides with long-standing historical legacies of colonialism, discrimination, and war. In such a context, it might be argued that states have particularly onerous obligations to allow asylum seeking if they have historically had a hand in the conditions that have led to persecution in the first place. Such responsibility might be relatively direct and contemporaneous. For example, Iraqi asylum seekers might have a good argument for seeking asylum specifically in the US and the UK, countries that invaded them illegally, precipitating a predictable spiral of violence and resulting persecutions.Footnote ⁴ Afghans could emphasize that they would not have been persecuted by the Taliban had it not been for the US’s sudden and ill-planned pull out.

One could also imagine that asylum seekers would have a particularly strong normative claim to seek asylum in a particular country if that country were a former colonial power whose legacy had created some of the very sources of persecution that led them to flee.Footnote ⁵ Indeed, it may be that persecution in one’s state is linked to longer-term colonial or imperial legacies that have stigmatized certain populations, fostered dictatorship, or made repression endemic. Under that light, for example, Belgium would have a particularly strong responsibility to welcome Rwandan asylum seekers suffering from continued ethnic legacies largely fomented by Belgium colonization (or, which is almost the same, such asylum seekers would have a particularly strong case to make their demands in Belgium).

What agency brings to the table of theories about state responsibilities is the ability to transform the relationship to asylum seekers as one of accountability and not simply of abstract cosmopolitan justice. Within this accountability framework, refugees’ de facto agency acts as a reminder, trigger, and, sometimes, pointed finger. Finally, the argument proposed therein for country choice may be particularly strong in the case of asylum seekers, but that does not nullify its broader viability for all would-be migrants based on a similar analysis of the imbrication of the state of origin and the host state. For example, the movement to theorize reparative justice obligations vis-à-vis climate refugees shows the potential of a remedial approach in further blurring the distinction between conventional refugees and “mere” migrants.

Conclusion: The Border Shapes the Refugee, but Can the Refugee Shape the Border?

In this chapter, I have argued that the division constructed by the border between refugees and migrants has the unfortunate effect, among others, of rendering invisible refugees’ migratory trajectories. This is largely out of a fear that emphasizing such trajectories might actually weaken refugees’ claim to being refugees. This may well be the case in practice, and one may understand the merely tactical need to tone down the extent to which asylum seekers may have strong preferences for seeking asylum in certain countries, lest they be confused with migrants. But I have also stressed that there is no incompatibility between being a refugee and a migrant if one sees asylum seekers’ strategies as complex demands not just for protection in the abstract but for protection in certain countries; and, in fact, not just for protection but, having been forced to leave their states, for the possibility of rich and meaningful lives in particular places in which they came to seek protection.

In effect, refugees may be shaped by the border to act like univocal refugees more than is reasonable or will do justice to their multifaceted lives. What is clear, however, is that refugees also consistently express, most notably through their decisions to leave and travel, visions that one normally associates with migrancy. The problem, however, is that there is no normative narrative to account for such preferences. If choice of country is not to be left to the chance of a quasi-Darwinian struggle (Shachar, Reference Shachar2020b), then under the standard narrative that provides no room for recognizing the agency of refugees, it is at best left to efforts by states to share the burden of asylum seeking in ways that have no logical connection to asylum seekers’ own preferences.

Generally speaking, then, the political cost of the “border within” might be understood as the cost of a culture of bad faith, a culture in which one needs to adopt identities that maximize one’s life prospects for purely instrumental reasons. One result of that culture of hypocrisy is the development of a heavy regulatory and police response to constantly discipline these categories and make sure they are not abused even as the border creates considerable pressures for them to be abused. The border thus acts as a self-fulfilling prophecy, producing the categories that are then internalized by migrants and asylum seekers, and also justifying constant policing to ensure that they continue to act as fix markers of identity. Another political cost of the border might be the erosion of solidarities between “mere migrants” and asylum seekers, even as an increasing body of evidence suggests the proximity of their fates (Scherr, Reference Scherr and Kury2018).

Contra a vision of the international refugee regime being merely and exclusively about protection, an emancipation from the relentless discipline of the border would highlight the case for taking into account asylum seekers’ agency about what country they want to obtain protection in. Indeed, in demanding asylum in this or that country, refugees manifest themselves as political agents of change making pointed demands to particular polities as well as the international system. As such, they also contribute to constantly redrawing and reinventing the border as a porous membrane that cannot prevent the insistent manifestation of political demands to circumvent it.