Refugee Responsibility Sharing or Responsibility Dumping?

Refugee Responsibility Sharing or Responsibility Dumping?

A silver lining of recent migration crises is increased reliance on responsibility-sharing arrangements in international actor responses. This new experience allows for evidence-based analysis of such arrangements. We distinguish between progressive arrangements—ones that shift responsibilities to more affluent, institutionally competent, and safer countries—and regressive arrangements that do the opposite and in fact constitute responsibility dumping.

Focusing on examples from the United Nations, the United States, and the European Union, we classify different responsibility-sharing arrangements as progressive or regressive based on four parameters: hosting commitments, monetary or equivalent contributions, multilateralism, and legally binding instruments. The analysis suggests that first safe country arrangements are presumptively regressive, while arrangements that include hosting or material commitments are presumptively progressive. The 2015 EU model and the 2021 U.S.-Afghanistan measures are particularly progressive models that ought to spread. We also highlight the importance of assessing refugee policy solutions with a view to broader, systemic implications for host countries and not just against an individual-focused legal threshold.


Migration crises in the last several years drove millions of international protection seekers to leave their homes in Africa, the Middle East, Central America, and South East Asia to seek refuge in other countries. A handful of developing countries have been disproportionately affected and host the vast majority of refugees.[1] These realities have fueled consideration of various arrangements designed to relieve the pressure on heavily burdened receiving states. “Responsibility sharing” has become a key concept in the policy debate around potential solutions that would alleviate the burden on receiving states while addressing the plight of international protection seekers.

The concept of responsibility or burden sharing is not new to refugee law and scholarship.[2] The preamble to the 1951 Refugee Convention, the cornerstone of the international refugee law regime, includes a general obligation to assist refugee receiving countries through international cooperation.[3] However, scholars have portrayed responsibility sharing as a largely theoretical, hortatory frame[4]—one that lacks concrete substance and requires development and translation “[f]rom Principle to Policy,” as Michael Doyle observed as recently as 2018.[5] Peter Schuck, who defended the concept of refugee responsibility sharing in an influential 1997 article, still maintained in his 2013 revisiting of the article that “[n]o strong norm of refugee burden-sharing currently exists in international law or practice.”[6] Rebecca Dowd and Jane McAdam similarly argued that “a mechanism to systematically, equitably and predictably allocate responsibilities between States at a global level has still not been agreed.”[7]

This Essay challenges these common perceptions among scholars and practitioners. Over the last several years, concrete responsibility-sharing arrangements have been developed and implemented, translating the concept from theory to practice. This allows us to move the refugee responsibility-sharing debate forward by mapping, analyzing, and assessing various responsibility-sharing models that are currently not well understood. The analysis should be of interest to both progressives who believe that responsibility sharing remains little more than an aspiration and to some conservatives who hold a minimalist view of what responsibility for international protection seekers entails.

The Essay delves into the details of responsibility-sharing policies and proposals considered or implemented by key actors: the United Nations, the United States, and the European Union. It highlights the legal devices that facilitate these policies, including first safe country requirements in domestic immigration laws, a growing number of major bilateral agreements shifting responsibility from one country to another,[8] and translation of abstract legal principles like the EU solidarity principle[9] into actionable, concrete responsibility-sharing formulae.[10]

Not all responsibility-sharing arrangements are adequately suited for the task they set out to accomplish. Some arrangements genuinely seek to establish responsibility sharing that helps relieve the burden on first receiving states while ameliorating the situation of international protection seekers. By contrast, other responsibility-sharing arrangements actually constitute responsibility dumping. Responsibility dumping mechanisms transfer responsibility for protection seekers from receiving countries to other countries that are not as wealthy, not as institutionally competent, or not as safe without adequate compensatory mechanisms.

In other words, responsibility-sharing arrangements may be highly progressive or highly regressive, depending on their details. We identify four parameters for assessing responsibility-sharing arrangements and consider how they apply in examples from the United Nations, the European Union, and the United States:

(1) protection seeker hosting commitments, meaning agreeing to take protection seekers in or declining to exercise a right to move migrants out of the host’s territory. As we discuss below, classic examples include the United States’ longstanding refugee resettlement program and Germany’s decision to allow over one million refugees to remain in its territory following the 2015 European refugee crisis;

(2) monetary component or equivalent non-monetary assistance and how it corresponds with estimated costs of protection seeker hosting and resettlement;[11]

(3) bilateral vs. multilateral arrangements, distinguishing multilateral arrangements, such as the EU solidarity measures, from bilateral arrangements, such as the Trump administration’s Asylum Cooperative Agreements (ACAs) with three Central American countries; and

(4) legally binding instruments, that is, whether an arrangement or instrument legally binds the parties. For instance, the UN Global Compact on refugees is entirely voluntary, while the EU 2015 emergency measures discussed at greater length below legally bind EU member states.

The first two parameters relate to the substance of the arrangements, while the other two relate to their form. All parameters may exist simultaneously within particular arrangements. For instance, monetary payments can be framed as binding legal obligations on each state or contributions to a voluntary multilateral mechanism. Multilateral mechanisms may be voluntary, legally binding, or consist of a mix of binding and non-binding obligations.

Based on these parameters, various responsibility-sharing arrangements can be placed on a scale.



On one end of the scale are maximally regressive arrangements. These are non-binding individual arrangements that entail no hosting or monetary commitments, ones that relocate international protection seekers to an unstable, less affluent country with weaker institutions. One example is the Trump administration’s ACAs with northern triangle countries. On the other end of the spectrum are maximally progressive arrangements. Those are multilateral, legally binding arrangements that entail significant hosting commitments and monetary commitments that correspond with realistic cost assessments of hosting each refugee. Such arrangements transfer international protection seekers to more stable, affluent countries with sounder and better-functioning asylum and migration institutions. As we will see, the EU 2015 emergency measures and the Biden administration’s efforts to resettle Afghan asylum seekers after the 2021 U.S. withdrawal from Afghanistan are examples of progressive mechanisms.[12]

What follows considers the details of concrete recent responsibility-sharing and responsibility-dumping arrangements, analyzes them based on these parameters, and places them in comparative context. Table 1 summarizes the analysis:


Table 1

Hosting Money Multilateral Binding
UN Global Compact ✓-* ✓-* X
U.S. Northern Triangle ACAs X X** X
U.S. Refugee Resettlement Program X X
U.S. 2021 Afghanistan Measures ✓-***
EU 2015 Emergency Measures
EU-Turkey 2016 Joint Statement X [?]
EU 2020 New Pact Proposal ✓-****
EU Border Policing Initiatives X [?]

* There is no specification of substantive contribution commitments. Member States are encouraged to contribute in any way they can.

** As we will see, while the agreements included payments from the United States to recipient third countries, the payments largely restored aid that the United States revoked to pressure those countries into signing the agreements.

*** The Biden administration has made substantial efforts to host Afghan asylum seekers who had helped the U.S. government presence in the country. But the crisis affects a much broader category of displaced Afghans whose plight has been left unaddressed. Moreover, as we will see, “parking” Afghans in transit hubs around the world before entry into the United States raises significant concerns.

**** The proposed instrument—an EU regulation—is legally binding, but Member States retain broad discretion in selecting their substantive commitments.


The Essay, then, offers several key contributions. First, it informs theoretical and legal discussions about responsibility sharing through an assessment of practice. It studies concrete arrangements and the legal devices that facilitate them in comparative context. It identifies progressive second-best responsibility-sharing models that ought to spread—particularly the EU 2015 solidarity mechanism and the 2021 U.S.-Afghanistan measures—while warning against the increased diffusion lately of highly regressive models that ought to be discouraged.[13] It shows that responsibility-sharing (and dumping) arrangements have expanded to include the broad and diverse category of asylum seekers, not only the smaller category of individuals already granted refugee status.[14] Second, the Essay argues that we should focus not only on duties to refugees, but also duties to host states. Third, the Essay shifts the focus from first receiving countries to the responsibilities and potential contributions of other stakeholders. Finally, the Essay lays a foundation for the development of international norms and perhaps—down the road—even binding legal obligations around responsibility-sharing arrangements.

I.  Shifting Perspective to Host States

The concepts of responsibility sharing—or burden sharing—permeate current policy discourse and implementation regarding seekers of international protection. It is invoked across jurisdictions and ideologies. As we will see, responsibility sharing—solidarity—is a foundational concept in EU asylum and migration policy. It is a cornerstone of the UN Global Compact on Refugees, and a key tenet of the African Union’s approach.[15]

Likewise, when the Trump administration defended its controversial immigration regulations before the U.S. Supreme Court, the administration emphasized that it “encourages foreign countries to ‘partner’ with the United States and to shoulder their share of the burdens of mass migration.”[16] When the Biden administration suspended the ACAs the Trump administration had concluded with three northern triangle countries, it underscored that “the United States is taking this action as efforts to establish a cooperative, mutually respectful approach to managing migration across the region begin.”[17] The United States, the administration stated, “will build on our strong relationships and support these governments’ efforts to address forced displacement without placing undue burden on them . . . .”

However, not all these invocations of responsibility sharing should be treated equally. Some of the policies promoted under this concept genuinely seek to ameliorate the situation of protections seekers and host states alike. By contrast, others seek to shift responsibility away from their proponents without any mitigating measures. How, then, should we distinguish between responsibility-sharing arrangements that should be encouraged as models for addressing asylum and migration challenges and ones that should be discouraged?

Lawyers tend to assess responsibility-sharing arrangements against minimalist, individualized international refugee law obligations. A responsibility-sharing arrangement is acceptable under this approach if it meets basic requirements, including the principle of non-refoulement, which prohibits states to turn individuals back to danger. At a minimum, compliance with this principle requires states to implement adequate asylum application mechanisms. Once an arrangement meets that legal threshold, this individual-centric approach is neutral on the components and larger impact of the arrangement.

For instance, in 2013, UNHCR issued a guidance on “bilateral and/or multilateral transfer arrangements of asylum seekers” that recognized the growing number of responsibility-sharing arrangements worldwide and proposed criteria for their assessment. The guidance was framed in binary terms: either a receiving country met certain legal and safety requirements, or it did not, in which case refugee and asylum seeker transfers should not take place. The guidance provided that the receiving state should be a party to the 1951 Refugee Convention and must guarantee that it would fully comply with relevant legal obligations, including individual assessment of asylum seekers, non-refoulement, fair process, and access to basic benefits.[18]

Similarly, the Trump administration defended requiring applicants for asylum in the United States to first apply in the first safe country of transit by arguing that the policy is fully consistent with international law and practice. The administration contended that the policy was no different than first safe country arrangements adopted by the EU and other liberal democracies.[19] The administration repeated this claim in a rule designed to implement the ACAs with Guatemala, Honduras, and El Salvador.[20]

Focusing on whether an arrangement meets a basic legal threshold, however, is presumptively insufficient. There is significant variation among potential arrangements above the legal threshold. A certain arrangement could meet minimal legal requirements but remain regressive because it entails transferring an international protection seeker from a stable, relatively affluent country to a weaker, less safe, and less affluent one. Generally, arrangements that require protection seekers to apply for status when they arrive in the first safe country are inherently regressive—first safe countries are nearly always developing countries suffering from significant resource and institutional constraints.

We therefore propose a different metric that goes beyond minimal legal requirements and abstract references to burden sharing. Our framework focuses instead on whether a given arrangement is, on balance, progressive or regressive. Our evidence-based analysis shifts the perspective from individual asylum seekers to systemic approaches that consider the challenges posed by mass influxes of asylum seekers from the perspective of host states.[21]

It is important to emphasize at the outset that we are concerned here with responsibility-sharing arrangements as second-best solutions. Many refugee law and human rights advocates and scholars would argue that states should not seek to relieve their own burden through sometimes dubious responsibility-sharing arrangements.[22] States, the argument goes, should open their borders to any individual seeking international protection, diligently process and assess their eligibility for asylum, and, if granted, provide adequate support. States should strictly follow the non-refoulement principle, prohibiting the return of international protection seekers to countries where they may face danger. Indeed, scholars and advocates have argued that responsibility- and burden-sharing arrangements collectivize and commodify international protection seekers instead of treating them as “human beings in peril.”[23]

These critiques certainly carry significant weight. We must recognize, however, that current realities fall well short of this idealized aspiration for the international refugee regime due to intractable structural and political reasons. It is therefore necessary to carefully assess second-best alternatives and distinguish them from regressive third-best “solutions” that should be rejected. As the work of Ayelet Shachar has demonstrated, the latter type of arrangements has quickly diffused across jurisdictions in recent years.[24] There is much work to be done in separating progressive second-best responsibility-sharing arrangements with policy advantages for both states and international protection seekers from significantly worse, regressive arrangements disguised as responsibility sharing that in fact constitute responsibility dumping.

II.  The UN Global Compact: Passing the Collective Collection Plate


Hosting Money Multilateral Binding
UN Global Compact ✓- ✓- X

The concepts of responsibility sharing and “international solidarity” are the primary guiding principles of the United Nations Global Compact on Refugees.[25] UNHCR drafted the Compact pursuant to the New York Declaration for Refugees and Migrants mandate, and it was endorsed by the General Assembly’s 2018 annual UNHCR resolution.[26]

The Compact “seeks to operationalize the principles of burden- and responsibility-sharing to better protect and assist refugees and support host countries and communities.”[27] In line with this principle, the Compact proposes an action plan to support refugees and “countries particularly affected by a large refugee movement, or a protracted refugee situation, through effective arrangements for burden- and responsibility-sharing.”[28] It envisions more equitable and predictable arrangements at the global, regional, and country level.

Similar to other UN documents of its kind, the Compact is laden with aspirational and voluntary language. It is not legally binding, and it is replete with state interest caveats. Nevertheless, the Compact has several practical components.

First, the Compact established the Global Refugee Forum (GRF), a high-level forum of UN member states and other stakeholders to be convened every four years beginning in 2019. The GRF was designed to serve as a platform for states and other stakeholders to coordinate and announce concrete pledges and contributions toward the realization of the Compact’s objectives. If you will, it is an exercise in passing the collective collection plate and encouraging states and other actors to contribute as much as they are willing and able to contribute. State GRF pledges may include “financial, material and technical assistance, resettlement places and complementary pathways for admission to third countries,” among others.[29]

The first GRF was in fact convened in late 2019. The convenors of the Forum reported more than 840 pledges and contributions by a variety of actors, including governments, NGOs, and even private sector entities.[30] Of these, roughly 250 pledges contained a financial component, including major monetary contributions by the World Bank Group and the Inter-American Development Bank ($4.7 billion and $1 billion, respectively).[31] Other GRF pledges of financial support exceeded $2 billion, and the private sector pledged an additional $250 million.[32] The convenors also commanded states’ commitment to developing better refugee integration policies within their own jurisdiction, as well as the establishment of an Asylum Capacity Support Group and the Global Academic Interdisciplinary Network to promote the objectives of the Global Compact.

Second, the Compact envisioned a new mechanism whereby host countries may request to activate “Support Platforms” (SPs) to address their specific challenges. Like the GRF, SPs are designed to help mobilize “financial, material and technical assistance, as well as resettlement and complementary pathways for admission to third countries . . . where applicable, drawing on Global Refugee Forum pledges.”[33] SPs appear to be intended primarily for large-scale or protracted refugee situations that overwhelm the response capacity of a host state. UNHCR has a key role in facilitating, coordinating, and overseeing them. Among the tools the Compact proposed for SPs to utilize is initiating “solidarity conferences” to assist host countries.[34]

Some scholars have criticized the Compact for eroding state obligations under international refugee law because it commodifies refugees and collectivizes what is in essence an individualized regime. They warn against the risk of states basing resettlement decisions on a refugee’s skills and earning capacity instead of their individual need and the risk they face in their country of origin.[35]

Other evaluations of the Compact are not quite as dire. For example, Alexander Betts argued that the Compact’s success in generating unprecedented global consensus around ideas of refugee-related responsibility sharing is itself an achievement, while recognizing that the Compact’s substance is actually fairly modest. He observed that the Compact’s “basic aim . . . is to get more resources into the system rather than to change the system.” In his view, the Compact essentially does three things: it cites all the actors that can contribute to responsibility sharing, identifies the areas in which they can contribute, and initiates new structures to facilitate and promote contributions like the GRF and SPs. Betts maintains that it is too soon to tell what the Compact’s actual practical effects will be and whether its responsibility-sharing provisions will be effective.[36]

Clearly, the Global Compact is not an ideal solution for the problem of uneven burden distribution among states in receiving the flows of migrants fleeing crisis. It is voluntary and underspecified. It does not require concrete monetary or equivalent contributions, and it relies heavily on stakeholders’ good will.

That being said, the Compact does offer second-best solutions that have shown some promise in encouraging a variety of stakeholders to contribute either monetary or other assistance to the collective global effort to address the refugee challenge. The 2019 GRF produced an arguably large number of contributions, including significant monetary contributions. The Compact acknowledges the need for responsibility sharing, creates collective processes to achieve greater buy-in from all relevant global players, establishes at least a minimal degree of accountability, and offers some support for states and regions facing heavy migration burdens. We therefore view the compact as a relatively progressive responsibility-sharing arrangement.

III.  The United States


Hosting Money Multilateral Binding
U.S. Northern Triangle ACAs X X X
U.S. Refugee Resettlement Program X X
U.S. 2021 Afghanistan Measures ✓-


Since the enactment of the Refugee Act of 1980, the United States has consistently been at the top of the world’s list of leading resettlement countries. According to UNHCR third country resettlement data, “In recent years, the United States has been the world’s top resettlement country, with Canada, Germany, the United Kingdom, Australia and the Nordic countries also providing a sizeable number of places annually.”[37]

The United States has remained at the top of UNHCR’s resettlement destination list despite restrictive policies put in place by the Trump administration,[38] discussed in greater detail below. However, under the Trump administration, the United States was no longer the top resettling country overall (that is, if we account for all resettlements, not just UNHCR-facilitated ones).[39] The next Section explains the Trump administration policies that contributed to this outcome.

Despite the shrinking of the U.S. resettlement program under the Trump administration, the United States maintained its status as the top donor to the UNHCR. As of December 2020, the United States rested atop the donor list, with a total contribution exceeding $4.7 billion—more than double the second-largest contribution of the European Union.[40] At the same time, in 2018, the Trump administration cut funding for the United Nations Relief and Works Agency for Palestine Refugees (UNRWA). In 2017, the U.S. contribution to UNRWA amounted to one third of the agency’s annual $1.1 billion budget.[41] In so doing, the Trump administration reversed a policy that had enjoyed bipartisan support since the creation of the agency seven decades earlier.[42]

A.  The Trump Cuts

The Refugee Act of 1980 authorizes the president to set an annual refugee admission ceiling. The caps represent an annual maximum, which the U.S. government is not required to reach. The Trump administration systematically cut refugee caps each year since it took office in 2017.[43] It also gutted the infrastructure of the United States Refugee Admissions Program (USRAP).[44]

The Biden administration has so far advanced a markedly different approach in rhetoric, if not in practice.[45] Candidate Biden vowed to “set the annual global refugee admissions cap to 125,000”—exceeding Obama administration numbers—“and seek to raise it over time commensurate with our responsibility, our values, and the unprecedented global need.”[46] Initially, after Biden was elected, the administration signaled it would retain the low Trump admission cap. Following public pressure, the Biden administration set the cap at 62,500 for the remainder of 2021.[47] It also raised the cap for 2022 to 125,000.[48] Crucially, the Biden administration expanded the resettlement program to assist Afghans in the aftermath of the U.S. withdrawal from Afghanistan. Section C examines these measures in greater depth.

Refugee resettlement programs like USRAP, which involve admission of individuals not already located in the resettling state, are inherently progressive responsibility-sharing arrangements. The resettling state commits to admitting and providing for refugees. However, the Trump administration’s policies and their impact on U.S. refugee admission highlight significant weaknesses in the statutory architecture that facilitates the USRAP. The Refugee Act delegates broad discretion to the president to set annual refugee admission caps. Refugee admission and its scope are therefore voluntary and discretionary. A different statutory scheme could have prescribed binding annual caps, which would be less susceptible to fluctuation based on the priorities and exigencies of a given administration. The discretionary nature of USRAP allows presidents who are not committed to refugee protection to undermine the program, as the Trump experience illustrates.

B.  Aid to Guatemala, El Salvador, and Honduras in the Framework of ACAs

In addition to cutting refugee resettlement caps in the United States, the Trump administration took significant steps to relocate asylum seekers and deflect applications to third countries in Central America through so-called third safe country agreements. Among other measures,[49] between 2019 and 2020, the Trump administration signed bilateral Asylum Cooperative Agreements, or ACAs, with Guatemala, El Salvador, and Honduras.[50] The only ACA the United States had at the time was a 2002 agreement with Canada.[51] Under these agreements, the three countries agreed to accept asylum seekers from the United States, process asylum applications, and ratchet up efforts to stem illegal immigration to the United States through their territory.

In return, the Trump administration promised these nations economic benefits. Importantly, however, the “benefits” came after the Trump administration’s March 2019 decision to cut all aid to the three countries over their alleged failure to address illegal migration to the United States transiting through their territory. These cuts amounted to a total of $615 million in assistance.[52]

After the three countries agreed to U.S. demands, in June 2019, the Trump administration restored a large portion of this aid. It resumed aid in the amount of $432 million in projects and grants that had previously been approved.[53] In April 2020, the Trump administration said it would make additional funding available to El Salvador, Guatemala, and Honduras. The administration attributed the move to increased cooperation from the three countries in combatting illegal migration to the United States. In June 2020, then-Secretary of State Pompeo announced $252 million in additional U.S. foreign assistance for El Salvador, Guatemala, and Honduras combined.[54]

In other words, the assistance provided to the three countries in return for their agreement to accept asylum seekers from the United States was little more than reinstatement of foreign aid previously revoked by the Trump administration in order to pressure them into entering these agreements. No real effort was made to ensure that Guatemala, Honduras, and El Salvador offered a safe environment for returned migrants, or that the countries would adequately process asylum applications. As many have observed, there are very good reasons to doubt that was the case.[55] This example therefore represents an extremely regressive arrangement. It cannot be viewed as a form of responsibility sharing. Rather, it is a striking example of responsibility dumping—diverting asylum seekers to less affluent, less institutionally competent, less safe countries without contributing funds or assistance to ensure their safety.

In defending other policies designed to deflect migrants and asylum seekers to Central American countries, the Trump administration asserted that those policies aligned with similar policies of other liberal democracies, including EU member states.[56] Such policies require seekers of international protection to apply for asylum in the first safe country through which they transit. As we will see in the next Part, the EU Dublin Regulations indeed enshrine a similar principle by requiring protection seekers to apply in the EU member state of first irregular entry. Put in context, however, it becomes clear that the Trump administration’s policies were far more regressive than their EU counterparts. The Dublin regulation deflects applicants to other EU member states, with resources, institutions, and safety conditions far superior to those of the Central American countries. What is more, the regressive effects of the Dublin regulations have been offset to a certain degree by other EU measures involving significant hosting commitments and monetary investment in refugees. No comparable arrangements exist in the United States that would offset the regressive effects of the third safe country agreements.

In any event, as it has done with the refugee resettlement program, the Biden administration reversed course on the issue of third safe country agreements. The Biden administration terminated all three agreements with Central American countries and opted for a different approach that would not place an “undue burden” on those countries.[57] The administration’s rhetoric implies that it accepts that a more equitable responsibility-sharing framework is needed to address the Central American migration crisis.

C.  2021 Afghanistan Measures

The United States’ withdrawal from Afghanistan in summer 2021 and the Taliban’s subsequent takeover of the country caused hundreds of thousands of Afghans threatened by the new regime to seek refuge outside the country. UNHCR has estimated that as many as half a million Afghans may have left the country by the end of 2021.[58] The current emergency exacerbated the pre-existing Afghan refugee crisis. The war-torn nation has produced 2.6 million registered refugees worldwide. 2.2 million Afghan refugees are registered in neighboring Iran and Pakistan alone. Another 3.5 million people are internally displaced.[59]

The Biden administration has taken several steps to address the situation. In the wake of the Taliban takeover, the United States coordinated an operation to airlift more than 124,000 people from Kabul’s international airport. As of September 26, however, only forty thousand people arrived in the United States.[60] Many evacuees were initially sent to third countries and U.S. military bases in Europe and the Middle East for security screening and processing. Nearly two dozen countries helped temporarily relocate at-risk Afghans, that is, those who assisted the United States and coalition forces, or others deemed particularly vulnerable. These countries include Albania, Bahrain, Canada, Colombia, Costa Rica, Chile, Ecuador, Guyana, India, Kuwait, Mexico, Netherlands, North Macedonia, Poland, Portugal, Qatar, Rwanda, Singapore, Slovakia, Uganda, and the United Arab Emirates.[61] The United States has mobilized the private sector as well.[62]

In addition, the Biden administration expanded pathways for Afghans to resettle in the United States. It gave Afghans who did not qualify for Special Immigrant Visas (SIVs)—designed for those who assisted U.S. forces in Iraq and Afghanistan—access to the USRAP.[63] Moreover, Congress added eight thousand SIV visa slots, increasing the existing 26,500 quota. It also appropriated over $1 billion in funding for the Defense and State Departments and other relevant agencies to address the Afghan refugee crisis. Congress allotted $100 million of this appropriation for assistance for Afghan refugees in neighboring countries.[64] The Biden administration has called on Congress to appropriate another $6.4 billion to “meet our commitments to Afghan allies and partners.”[65]

Despite these efforts, key legal and policy complexities remain unaddressed. For instance, thousands of Afghans were admitted to the United States through the humanitarian parole authority, which allows temporary admission without required documentation. These individuals are not immediately eligible for work authorization and permanent status.[66] Furthermore, the number of SIV visas available is exceedingly low compared to UNHCR’s projection of five hundred thousand Afghan refugees by the end of 2021. The status of those currently in transit hubs around the globe and on U.S. military bases remains uncertain, and questions arise as to the adequacy of their housing conditions.[67] Additionally, the United States has focused its efforts on U.S. allies and those who assisted its presence in Afghanistan. But the United States arguably bears responsibility for a far broader category of Afghans because of its prolonged presence in the country.[68]

Nevertheless, the United States’ response to the Afghan refugee crisis encapsules impressive progressive elements, on a scale rarely seen in its refugee policy. The response includes significant resettlement commitments in the United States, a substantial allocation of funds, a global effort to coordinate responsibility sharing for the crisis among many countries, as well as facilitating legislation that created a targeted resettlement priority program. On balance, then, the U.S.-Afghanistan measures sit close to the progressive end of our scale.

IV.  The European Union


Hosting Monetary Multilateral Binding
EU 2015 Emergency Measures
EU-Turkey 2016 Joint Statement X [?]
EU 2020 New Pact Proposal ✓-**
EU Border Policing Initiatives X [?]


The EU is perhaps the global actor that has made the most progress toward a binding migration and asylum responsibility-sharing (or, in EU parlance, “solidarity”) mechanism. In addition, the EU has advanced a patchwork of other policies and instruments to address increased migration flows around the 2015 crisis. Such policies include the 2016 Joint EU-Turkey Statement and arrangements designed to increase enforcement against unlawful migration to Europe via the Mediterranean Sea.

Assessing this set of EU policy initiatives reveals a mixed record. The EU post-2015 solidarity model—the 2015 emergency measures and the New Pact proposal—includes highly progressive elements. Yet other arrangements, like increased border policing and cooperation with third countries to prevent migrants from reaching the EU, have produced concerning regressive effects.

Still, the EU experience gives reasons for optimism about the trajectory of refugee responsibility sharing. We have witnessed a gradual translation of an abstract principle of EU asylum and migration law—the solidarity principle enshrined in article 80 of the Treaty on the Functioning of the European Union—into actionable obligations for each state. The extent of the obligations is determined by concrete, progressive allocation formulae composed of non-negotiated criteria. An external actor—the EU Commission—centrally allocates obligations. The allocation criteria, mainly member states’ size and wealth, are objective. The system benefits small, less powerful EU member states that would have been likely to bow to more powerful member states had obligations been negotiated ad hoc for every migration crisis.[69] The EU model also expands on U.S. efforts to relocate refugees from Indochina in the 1980s by addressing the much larger and more diverse category of asylum seekers.[70] Therefore, although the current state of the EU asylum and migration system leaves much to be desired and is constantly the subject of reform efforts, it represents progress in operationalizing responsibility sharing.[71]

What follows considers and evaluates the evolving EU solidarity mechanism. We address the 2015 emergency measures, the Dublin IV proposal, and the New Pact proposal. We find that those initiatives are largely progressive responsibility-sharing arrangements. We then turn to the EU’s increased border policing initiatives, focusing particularly on the 2016 EU-Turkey joint statement aiming to return asylum seekers from Greece to Turkey. We find that the agreement includes significant progressive elements despite heavy criticism it has met.

A.  The Evolving EU Solidarity Mechanism

An aspiration for better responsibility sharing among EU member states with respect to migrants and asylum seekers has led to protracted efforts to reform the Common European Asylum System (CEAS), which was originally established in 1999. The refugee crises of 2015 and the pressure on member states with external borders—those that faced the greatest influxes of migrants and asylum seekers, particularly Greece and Italy—have added urgency to these efforts.

1.  The 2015 EU Emergency Measures

In fall 2015, the EU passed a set of emergency measures to assist Greece and Italy. At the time, an influx of migrants and asylum seekers, a result of the civil war in Syria and the wars in Afghanistan and Iraq, overwhelmed the two countries. By December 2015, more than 911,000 refugees and migrants had arrived on European soil. The main migration routes extended from Libya to Italy through the Mediterranean Sea and from Turkey to Greece and further north into Europe.[72]

The EU approach went far beyond passing the collection plate and soliciting voluntary contributions. EU Council Decision 2015/1601 ordered the relocation of 120,000 seekers of international protection from Greece and Italy to other member states by assigning mandatory relocation quotas to each member state.[73] This added to a previous measure that required the relocation of an additional forty thousand individuals.[74] The EU calculated the number of individuals it required each member state to relocate based on a distribution formula not included in the final version of the decision. The formula did, however, appear in the EU Commission’s proposed draft decision.[75] It accounted for population size, GDP, the relative number of asylum applications fielded by each member state in the preceding four years, and unemployment rate.

In addition, Council Decision 2015/1601 provided for EU payments to the member states receiving solidarity assistance as well as payments to member states accepting re-located applicants for international protection. Recipient states received a lump sum of €6,000 per individual, while Greece and Italy received €500 for each individual relocated from their territories. The EU Asylum, Migration and Integration Fund (AMIF) disbursed a total of €225.6 million for all individuals relocated under this scheme.[76]

The quota system provoked heated political opposition from certain member states. Hungary and Slovakia, supported by Poland, challenged the measures at the Court of Justice of the European Union (CJEU). These states disputed the quota system and argued that they should not be required to accept any relocated international protection seekers. The Court rejected these claims and affirmed the measures.[77]

Even so, the implementation of the 2015 emergency measures was only partial. Under the measures, 34,705 protection seekers out of the net target of 98,256 were relocated.[78] Performance and compliance varied among member states. Many member states failed to meet their relocation quotas. Others declined to accept any relocated individuals at all. According to a 2019 EU Court of Auditors Report, although it was “relevant,” the EU “[t]emporary emergency relocation schemes did not reach their targets and main objective.”[79]

These data, however, tell an incomplete story. Certain countries that have been leading recipients of international protection seekers in 2015 independently of the EU emergency measures complied particularly poorly with the EU emergency relocation scheme. For instance, Germany was the top recipient of asylum seekers in 2015, receiving as many as 890,000 applications. This was pursuant to Chancellor Merkel’s decision that year to accept over one million refugees into the country.[80] Nevertheless, Germany only relocated 1,349 of its assigned 27,536 individuals (4.9 percent) under the EU emergency measures, and features among the bottom relocators. Germany’s limited compliance with the EU relocation scheme may be attributed to the bureaucratic backlog created in the German system due to the influx of applications for international protection it agreed to process voluntarily. As of 2017, the country had 200,000 to 250,000 pending cases.[81] This example illustrates that a state voluntarily declining to exercise its right to turn migrants away could prove significant in providing sustainable solutions for refugees.

In sum, both the EU 2015 emergency measures and Germany’s voluntary acceptance of international protection seekers represent important—albeit far from perfect—advances where it comes to responsibility sharing. The German case is an example of a clearly progressive responsibility-sharing measure. It involves asylum seeker hosting on a large scale while assuming related costs, and it entails the transfer of asylum seekers from less affluent and less institutionally competent EU member states to a wealthy state with a robust legal system and government bureaucracy. The shortfall of the German example for purposes of our assessment is its voluntary and unilateral nature.

Likewise, the EU emergency measures translated the solidarity and responsibility sharing concept into concrete, binding allocation formulae to determine each member state’s hosting obligations. The political backlash this approach has received, and the implementation gaps in the execution of the 2015 measures, indicate its novelty and progressivity. Recalling our parameters, the measures entailed both hosting obligations for member states and a monetary commitment by the EU to defray the costs of protection seeker transfers. As arrangements enshrined in an EU Council Decision, the measures were both multilateral and legally binding on EU member states. The EU emergency measures are therefore close to the progressive end of our scale.

2.  The 2016 Proposal to Reform the Dublin Regulations

The refugee crisis of 2015 and subsequent implementation struggles led to efforts to advance more permanent reforms of the EU responsibility sharing and solidarity mechanism. In 2016, the European Commission proposed to reform the Dublin Regulations—the key EU legal instrument governing the handling of applications for international protection.[82] The Dublin Regulations outline the “criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third country national or a stateless person.” Since the main criterion in previous versions of the regulations is the first point of irregular entry, the existing regime exacerbated the burden on external-border member states like Greece and Italy.

The 2016 reform proposal (Dublin IV) aimed to “[e]nsure fair sharing of responsibilities between Member States by complementing the current system with a corrective allocation mechanism” in cases of “disproportionate pressure.” Under that system, there would be an automatic threshold above which a member state would be deemed to be handling a disproportionate number of applications for international protection. A member state’s size and wealth determine that threshold. Once the threshold is reached, all subsequent applicants in the state would be relocated across the EU until the number of applicants in the state under pressure stabilizes.

The Dublin IV proposal reduced the number of national parameters to be accounted for in the formula used to determine how many asylum seekers each member state would be required to take in. It simplified the formula compared to the one used in 2015 such that Dublin IV only accounted for population size and GDP. In addition, the proposal introduced a significant alternative member states could opt for if they wanted to avoid relocating international protection seekers to their territory: the option to pay a “solidarity contribution” of €250,000 per applicant otherwise eligible for relocation to their jurisdiction under the fairness mechanism. That number is considerably high compared to the spectrum of available refugee resettlement cost assessments.[83] This suggests that the number did not only reflect an assessment of lifetime resettlement costs, but also that it incorporated a penalty component as a sanction on member states that decline to accept their share of relocated individuals.

In any event, the EU never adopted the Dublin IV proposal. Instead, in September 2020, the EU Commission announced that it was withdrawing the Dublin IV proposal and replacing it with a comprehensive New Pact on Migration and Asylum.[84] The New Pact consists of several elements, including the Asylum and Migration Management Regulation, a broad instrument for common asylum and migration management.[85]

3.  The New Pact Asylum and Migration Management Regulation Solidarity Mechanism

The New Pact Asylum and Migration Regulation outlines a new solidarity mechanism to address migration in pressure situations. Unlike the previous incarnations of the formulae for burden allocation among member states, it offers member states a flexible set of options instead of mandatory relocation quotas or pre-set payments. It envisions two kinds of solidarity contributions that member states will be under the obligation to provide: (1) relocation or return sponsorship; and (2) contributing to member state capacity building in the fields of asylum, reception and return and “in the external dimension.” Return sponsorship entails helping the member state entitled to solidarity assistance to repatriate and reintegrate third-state nationals present in the benefiting member state unlawfully, including through facilitating negotiations with third states. The scope of member state obligations will be calculated through a formula accounting for 50 percent GDP and 50 percent population. Member states would receive a €10,000 payment for each asylum seeker relocated to their territory.[86]

The New Pact Regulation proposal represents, in many respects, a step back from the previous iterations of the EU solidarity mechanism discussed above. It abandoned the method of imposing binding, specific relocation quotas on member states. It also left out the significant €250,000 payment obligation on member states for every individual they are required to but decline to host. Instead, it introduces a flexible, watered-down framework that allows member states to satisfy their obligations via an ill-defined menu of substitutes for monetary payments or hosting. Therefore, this proposed arrangement is not as progressive as the 2015 emergency measures or the defunct Dublin proposal.

Nonetheless, the New Pact Regulation proposal remains a multilateral binding legal instrument that enshrines a solidarity mechanism and includes quantifiable obligations member states would be subject to. Therefore, on balance, it still constitutes a progressive arrangement, albeit one that is closer to the regressive end of the spectrum than the earlier EU instruments and proposals. Michael Doyle, Janine Prantl, and Mark Wood even argued in their contribution to this issue that the New Pact’s “à la carte” approach to responsibility sharing is “a much more attractive alternative” than a binding quota system.[87] It creates flexibility and economic incentives for EU member states to assume responsibility for international protection seekers.[88]

B.  Increased Migration Border Policing

The relatively generous proposed responsibility-sharing arrangements in the EU evolving migration solidarity mechanism only tell one side of the story of EU efforts to address influxes of migrants and seekers of international protection. The EU and individual member states have been accused repeatedly of attempting to shirk their obligations under the international refugee regime. The EU has done so by working with third parties to prevent international protection seekers from reaching its shores in the first place (call it access blocking),[89] relocating international protection seekers to third countries that provide them inadequate conditions and protection (inadequate relocation), or pursuing readmission into states of origin without appropriate safeguards (inadequate readmission).[90]

Examples of access blocking include “measures such as special zones for policing migrants and asylum seekers established within the territory of another state, as well as incentivizing or coercing other states to counter unauthorized migration through enhanced patrols.”[91]

A possible example of inadequate relocation that has garnered a fair amount of attention and criticism is the 2016 agreement between the EU and Turkey.[92] Under this agreement, Turkey agreed to readmit migrants who entered Greece illegally through its territory in exchange for significant financial EU support for migrants in Turkey and other benefits. The agreement provides that “[a]ll new irregular migrants crossing from Turkey into Greek islands as from 20 March 2016 will be returned to Turkey . . . in full accordance with EU and international law, thus excluding any kind of collective expulsion.”[93]

The EU, in turn, agreed to resettle one Syrian refugee from Turkey for each Syrian returned to Turkey under the agreement. The EU monetary commitment under the deal amounts to roughly €6 billion to date, allocated for refugee support in Turkey. According to the agreement, return from Greece to Turkey is conditioned upon a negative asylum decision if an application is filed. Despite heavy criticism,[94] as well as implementation problems due to case backlogs in the Greek asylum system,[95] the EU has touted this agreement a success—highlighting significant reductions in irregular crossings from Turkey to Greece—and a model for future agreements with other countries.[96]

Examples of alleged inadequate readmission abound as well. Such was the readmission effort of the AU-EU-UN Tripartite Taskforce on the Situation of Stranded Migrant and Refugees in Libya. Between 2017 and 2019, this taskforce’s operations led to over forty-eight thousand ostensibly voluntary returns of migrants stranded or abandoned along migration routes in Libya back to their home countries.[97] More broadly, the EU has been accused of entering into readmission agreements with third countries through coercion and promises of various benefits without putting in place sufficient safeguards to protect returned international protection seekers.[98]

The access blocking and inadequate readmission examples are clearly regressive. They do not involve any hosting commitments on the part of the EU. Quite the opposite: they are aimed at preventing international protection seekers and other migrants from ever reaching the EU. The result is that protection seekers remain in or are returned to unsafe situations where they are unlikely to receive adequate conditions and protection. While these examples involve monetary investment by the EU in the form of incentives to transit countries and increased policing costs, those investments and economic incentives are not intended to benefit the protection seekers. Rather, they are used as political leverage on the states engaging in cooperation with the EU to block migration into its territory. Except for the increased policing initiatives, many of these examples are informal, non-legally binding arrangements. While they could be characterized as multilateral arrangements in some cases, such as the Libya initiative, they do not encourage buy-in from otherwise uninterested actors in a way that benefits protection seekers.

Assessing relocation to “safe” third countries, as evidenced in the EU-Turkey joint statement, is complex. The agreement has both progressive and regressive elements. On the one hand, the EU monetary contribution to improve conditions for migrants in Turkey is significant, especially considering the relatively small number of migrants returned to Turkey under the deal to date. Compared to the 540,000 migrants that arrived on the Greek islands in the first ten months of 2015, only 1,370 were returned to Turkey as of October 2017.[99] The EU commitment of €6 billion in monetary assistance to Turkey, earmarked for migrant support, is significant and envied by leaders of countries that receive far smaller international contributions to help address their own refugee crises.[100] The amount of funding per refugee is startling if we only consider the relatively small number of individuals transferred under the agreement, and it remains considerable even if we account for the total number of Syrian refugees currently in Turkey, approaching four million.[101]

In addition, the agreement includes a critical safeguard in conditioning returns from Greece to Turkey on a negative outcome of an application for asylum in Greece. Notably, the agreement also committed the EU to resettling Syrian refugees from Turkey, although that commitment largely built on existing member state obligations. Finally, at least for some international protection seekers, a solution that allows them to remain in the region instead of relocating to an EU member state might be preferable. Research on mobility bias has shown that migrants at times prefer to stay within local, national, or regional boundaries for cultural, religious, or family reasons.[102]

Nevertheless, translating the agreement from theory to practice has raised significant concerns. First, human rights advocates have focused on the untenable humanitarian conditions the deal created in so-called “hot spots” on the Greek islands—essentially, detention centers where irregular migrants entering from Turkey were held pending return to Turkey or review of their application for asylum by Greece. Second, the deal forced migrants to seek other, more perilous routes into Europe and to accept greater levels of risk. Third, many have questioned whether Turkey can reasonably be deemed a safe third country. Researchers and activists have argued that Turkey does not provide a real option for migrants to apply for asylum, that it often returns refugees to even more dangerous countries, and that it fails to provide adequate and safe conditions for migrants and international protection seekers.[103] In light of these concerns, some have argued that the EU willfully relied on an “untrue . . . premise that Turkey is a safe country for refugees and asylum-seekers.”[104]

On balance, then, the assessment of the EU-Turkey agreement reveals a mixed record. The arrangement is regressive in that it transfers migrants from an EU member state to a less affluent, less safe third country. But the agreement also contains progressive elements: the EU took on certain hosting commitments, and its economic contribution toward ameliorating conditions for migrants in Turkey was substantial under many available benchmarks.

V.  Conclusion: Progressive and Regressive Responsibility-Sharing Arrangements

Responsibility-sharing arrangements cannot be assessed as a single category. Whether they genuinely facilitate responsibility sharing or constitute responsibility dumping depends on their concrete elements. The Essay explored responsibility sharing from the perspective of host states. We analyzed specific policies in reference to a scale ranging from maximally regressive arrangements to maximally progressive arrangements based on four parameters: hosting commitments, monetary or equivalent contribution, multilateralism, and legally binding instruments. The analysis illustrates that the evolving EU migration solidarity mechanism and the UN Global Compact are closer to the progressive end of the scale. Other EU policies designed to deflect migrants are clearly regressive. The EU-Turkey agreement reveals a mixed record, closer to the regressive end of the scale. Likewise, the downsizing of the U.S. resettlement program under the Trump administration and the so-called “safe” third country agreements with Central American countries are highly regressive.

We argue that it is not enough to evaluate responsibility sharing against a legal minimum of basic individual asylum seeker rights. There is considerable variation among arrangements that could be said to meet the legal threshold. It is necessary to distinguish between progressive and regressive arrangements, and we propose presumptions for making these distinctions. Specifically, resettlement commitments or equivalent material contributions are presumptively progressive, while first safe country requirements and agreements are presumptively regressive.

As we have seen, many countries have taken on resettlement and material support commitments in situations that did not trigger their direct legal responsibility under international refugee law because those being resettled or supported through other contributions were not on their territory. The view that hosting commitments are presumptively progressive should not be controversial. We recognize the human rights community’s skepticism about monetary commitments as substitutes for hosting,[105] but experience shows that such commitments can go a long way toward ameliorating the situation of first receiving countries and international protection seekers alike.

The regressive nature of first safe country arrangements is a function of the geopolitical realities of migration. Frontier receiving countries tend to be developing countries that face their own resource constraints and security challenges. Deflecting protection seekers to such countries would generally involve transfer to less affluent, less safe, and less institutionally competent countries. This outcome should be deemed regressive even if basic legal requirements are met. Here we depart from the approach that the 2013 UNHCR guidance exemplifies, which would be agnostic about such arrangements.

Furthermore, the analysis of concrete arrangements illustrates that it is important to approach responsibility-sharing arrangements from the perspective of host states, not just the individual perspective of international protection seekers. Lawyers tend to focus on the latter, but that focus overlooks important aspects of mass refugee crises. An approach that accounts for the perspective of host states also invites deeper thinking about equitable distribution of responsibilities to stakeholders beyond receiving countries, as well as potential obligations that developed countries might have toward developing countries disproportionally affected by influxes of refugees.[106]

Finally, it is important to pay attention to responsibility sharing norm diffusion. Diffusion of progressive arrangements across different jurisdictions is arguably desirable, while proliferation of regressive arrangements such as flawed safe third country agreements and tactics for externalizing treatment of international protection seekers should be discouraged. We have seen an increase in the incidence of such regressive policies in the United States and the EU. It is necessary to encourage the diffusion of progressive arrangements like the EU solidarity mechanism to offset this trend.

These principles and insights should inform norm development regarding refugee responsibility-sharing mechanisms and guide related domestic policy. As we begin to consider concrete norm development in this area, it is essential that our thinking be grounded not only in theory, but also in evidence about the operation and effects of different responsibility-sharing arrangements.


Shortly before the publication of this issue, Russia invaded Ukraine and sparked yet another massive displacement. As of this writing, UNHCR estimated that over 2.3 million refugees fled Ukraine to neighboring countries.[107] The numbers keep rising. This time, however, the international response was swift and comprehensive.

The international response to the Ukraine refugee crisis is the best contemporary example of genuine responsibility sharing. Although commentators have rightly criticized the disparity between the European response to Ukraine and other recent migration flows involving Middle Eastern asylum seekers, and highlighted instances of racial discrimination at the border,[108] the positive features of the response model are worth analyzing—and replicating.

Once the Russian invasion began, neighboring European nations opened their borders to Ukrainians and many others living in Ukraine. Complementing these efforts at the EU level, the European Council granted Ukrainian nationals and others the right to live, work, and receive benefits in any EU country they choose, for at least one year and likely longer.[109] For the first time, the Council decided to activate the EU Temporary Protection Directive, a 2001 directive developed in response to the war in the Balkans to deal with mass influxes of refugees fleeing conflict.[110]

Protection seekers have never before enjoyed this degree of choice over the country where they will live and work. Ukrainians are now able to “choose the Member State in which they want to enjoy the rights attached to temporary protection and to join their family and friends across the significant diaspora networks that currently exist across the Union, rather than needing to apply for protection and stay in the first EU state they reach.”[111] The temporary protection directive, and the council decision implementing it, are also generous in their definition of a family. The definition includes not only spouses and minor children, but also other close relatives living with and dependent on the sponsor. These rights apply automatically and across the EU to Ukrainians and to persons under international protection (such as refugees) in Ukraine. EU member states have discretion on whether to extend these protections to other residents of Ukraine, and many will.

Moreover, the decision grants protection to those who will flee the Ukraine in the future, not only to those who already fled. Importantly, it removes procedural hurdles that could make it harder for Ukrainians to gain protection individually in EU member states.

Poland and Hungary supported these extraordinary measures and the council’s decision was unanimous. This is surprising considering that, in 2015, Poland and Hungary led the opposition to the EU emergency measures adopted to address the mass refugee crisis produced by conflicts in the Middle East.[112] But with Ukraine, these countries were sympathetic to robust steps at both the national and EU level. Popular sympathy for the Ukrainians, and the specter of further Russian westward expansion were likely key contributing factors. Whatever the ironies, an EU-wide response allows real choice and significant benefits to protection seekers.

The EU’s Ukraine measures go much further than any of the progressive responsibility-sharing arrangements that this Essay considers. The plan applies to all Ukrainians seeking protection, not just recognized refugees, and to their family members broadly defined. It also applies, on the same terms, to stateless persons and refugees from Ukraine. Individual EU member states are extending these protections to permanent residents of Ukraine. The measures provide for substantial hosting commitments and generous funding. Ukrainians will not only be permitted to remain in the EU. They will also gain access to work, education, health, and other benefits.

The European response to the Ukrainian refugee crisis is therefore at the very progressive end of our scale. Every one of the four elements of progressivity in our framework is met: there are significant hosting commitments, significant monetary commitments, a multilateral framework, and a legally binding instrument. Indeed, the EU Council allowed Ukrainians to select the country they will live and work in, a practice so novel we did not put it in our initial typology. The Ukraine response should be the model for handling future mass displacement situations.

Annex I: Assessing Refugee Hosting Costs

Assessing refugee resettlement costs is complex. Costs vary greatly depending on the resettling state, cost of living, division of responsibility between central and local government, benefit packages offered, transit costs, and the like. Although we cannot be confident what the exact numbers are, some generalizations based on available studies are nevertheless appropriate. Generally, assessments distinguish between immediate transfer costs and resettlement costs over a refugee’s lifetime. The first year appears to be the most expensive and costs decline over time.

For example, assessments of immediate costs of refugee resettlement in the United States are around $15,000.[113] A 2017 OECD report on the cost of integrating refugees noted that “[o]n average across the main reception countries in Europe, the cost for processing and accommodating asylum seekers is estimated around €10000 per application for the first year.[114] This figure can be significantly higher if integration support is already provided during the asylum phase . . . For the countries most affected by the recent refugee surge, the costs can thus be quite high. For example, Germany, which received as many as 900000 asylum seekers in 2015 . . . spent €16 billion (0.5% of GDP) on its migrants in that year. Sweden, which received 163000 asylum seekers in 2015 (the highest per capita ratio ever registered in the OECD at 1.6% of total population), spent €6 billion (1.35% of GDP).”[115] In the European Union, as we have seen, the EU cited a lump sum of €6,000 in emergency resettlement measures to pay for the initial transfer of an international protection seeker.

Costs are typically expected to be significantly lower in less affluent countries.[116] For example, the UN High Commissioner for Refugees has requested $1,057 per year to support each Syrian refugee in neighboring countries.[117] A study about international aid and refugee policy in Jordan states that the Jordanian government spent an estimated $8.6 billion in direct costs for hosting Syrian refugees between the onset of the Syrian civil war in 2011 and 2016. Syrians make up 6.8 to 13 percent of Jordan’s 9.6 million residents. The average cost per refugee therefore ranges between roughly $13,200 and $6,900 over five years.[118]

Addressing Colombia’s adoption of an open-door migration policy and reception of two million Venezuelan refugees, Colombia’s president asserted that his country spends upward of $1 billion a year on Venezuelan migrants. This translates to an annual cost of roughly $500 per migrant. He noted that international participation in these costs has been significantly lower than international participation in other recent refugee crises: “What has been pledged and disbursed in the case of Syria is more than $3,000 per migrant. When we look at the South Sudan migration crisis, we’re talking about more than $1,600 [per migrant]. And when we think about the [Venezuelan] crisis, it has been barely $316.”[119]

Cost assessments over the span of a lifetime vary greatly and tend to be influenced by overall approaches to immigration. Immigration skeptics tend to provide higher assessments, while progressives have put forward assessments as low as zero or argued that governments may even see economic gains from resettling refugees.

For instance, Richwine, Camarota, and Zeigler of the Center on Immigration Studies have estimated that “the average refugee’s lifetime fiscal cost, expressed as a net present value, is $60,000, with those entering as adults (ages 25 to 64) costing $133,000 each.”[120]

By contrast, Evans and Fitzgerald estimated in an oft-cited National Bureau of Economic Research study that the tax income refugees generate in the long run offsets and even exceeds the cost of resettlement: “On average, it costs about $15,000 to help settle a refugee, including both initial background checks as well as job and English training once they arrive. As refugees are also immediately eligible for welfare assistance and Medicaid, the government spends approximately $92,000 in governmental assistance for the first 20 years each refugee spends in the US. Over the same time, refugees pay an average of $129,000 in taxes — netting the government approximately $21,000 more than it spends.”[121] Similarly, a 2020 UNHCR report concludes, based on a survey of previous studies, that “the fiscal costs of refugee resettlement appear to be outweighed by the economic contributions of resettled refugees,” although it recognizes that it might take years to decades to offset resettlement costs.[122] A 2019 study focusing on the EU modeled the potential resettlement cost of all the refugees in Asia and Africa, finding that the average annual lifetime cost would not exceed 0.6 percent of the EU’s GDP.[123]

These assessments allow us to gauge the order of magnitude of resettling refugees (in USD) costs, distinguishing between immediate costs and resettlement costs over a lifetime. Immediate costs range from several hundred to the low thousands in developing countries. In developed countries, assessments are around $15,000. Costs over a lifetime are harder to measure because of the multiple parameters that go into these assessments. In developed countries, assessments range from zero (considering migrants’ contribution over a lifetime through work and taxes) to tens of thousands (USD). Several assessments conclude that resettling refugees leaves states with a net gain in the long run, considering refugee’s contributions to the workforce and the taxes they pay.[124]

Annex II: The Evolution of the EU Solidarity Contribution Formulae

  1. 2015 EU Emergency Measures to Assist Greece and Italy

The following formula was included in the EU Commission’s draft Council Decision proposal.[125] It did not, however, appear in the final version of the Council Decision.[126] The Decision only included two annexes (one for Greece and one for Italy) that outlined mandatory refugee acceptance quotas by member states, calculated based on the formula included in the Commission’s proposal:

“The proposed distribution key should be based on a) the size of the population (40% weighting), b) the total of the GDP (40% weighting), c) the average number of asylum applications per one million inhabitants over the period 2010-2014 (10% weighting, with a 30% cap of the population and GDP effect on the key, to avoid disproportionate effects of that criterion on the overall distribution) and d) the unemployment rate (10% weighting, with a 30% cap of the population and GDP effect on the key, to avoid disproportionate effects of that criterion on the overall distribution). . . .” [127]

  1. 2016 Dublin IV Regulation Reform Proposal

The Dublin IV reform proposal,[128] now defunct, reduced the number of national parameters to be accounted for in the allocation formula and simplified it such that it only accounted for population size and GDP.[129] In addition, the proposal introduced a significant alternative member states could opt for if they wanted to avoid relocating international protection seekers to their territory: the option to pay another member state a significant sum for every allocated applicant diverted to a different member state. EU member states would be given the option to opt out of the applicant reallocation scheme in exchange for a “solidarity contribution” of €250,000 per applicant otherwise eligible for reallocation to their jurisdiction under the solidarity mechanism. The payment recipients would be the member states that replace it as receiving states.

  1. The Proposed 2020 “New Pact” Comprehensive Draft Asylum and Migration Regulation

The proposed Comprehensive Draft Asylum and Migration Regulation[130] outlines a new solidarity mechanism to address migration in pressure situations. Unlike the previous incarnations of the allocating formulae, it offers member states a flexible set of options instead of mandatory relocation quotas or pre-set payments. It envisions two kinds of solidarity contributions that member states will be under the obligation to provide:

(1) relocation or return sponsorship

(2) contributing to member state capacity building in the fields of asylum, reception and return and “in the external dimension.”

Return sponsorship entails supporting the member state entitled to solidarity assistance in efforts to repatriate and reintegrate third-state nationals present in the benefiting member state unlawfully, including through facilitating negotiations with third states.

Under Article 54 of the draft Regulation, the scope of member state obligations will be calculated through a formula accounting for 50 percent GDP and 50 percent population. The proposed Regulation states that the distribution key will include the share of the benefitting member state to ensure that all member states contribute to fair responsibility sharing.

Katerina Linos: Professor of Law, and Co-Director, Miller Center for Global Challenges and the Law, UC Berkeley.

Elena Chachko: Rappaport Fellow, Harvard Law School, and Fellow, Miller Center for Global Challenges and the Law, UC Berkeley. We are grateful to the Miller Center, DAAD, and the Carnegie Foundation for their generous support, and to E. Tendayi Achiume, Seth Davis, Ryan Goodman, Kate Jastram, Jaya Ramji-Nogales, participants of the NYU Hauser Colloquium, and participants of the California Law Review Symposium on Sharing Responsibility for Refugees for invaluable comments and discussion. We thank Vanessa Rivas Bernady for excellent research assistance.

           [1].     As of mid-2020, 86 percent of forcibly displaced people worldwide were hosted in developing countries. 39 percent were hosted in just five receiving countries: Turkey, Colombia, Pakistan, Uganda, and Germany. Refugee Data Finder, UNCHR, []. According to UNCHR data, Turkey is the world’s top host, with 3.6 million forcibly displaced people. Colombia hosts 1.8 million, Pakistan and Uganda host 1.4 million each, and Germany hosts 1.1 million. Id.

           [2].     See e.g., Peter H. Schuck, Refugee Burden-Sharing: A Modest Proposal, 22 Yale J. Int’l L. 243, 243 (1997) [hereinafter Schuck (1997)]; Peter H. Schuck, Refugee Burden-Sharing: A Modest Proposal, Fifteen Years Later, Yale L. Sch. 1 (2013) [hereinafter Schuck (2013)]. See also Joseph Blocher & Mitu Gulati, Competing for Refugees: A Market-Based Solution to a Humanitarian Crisis, 48 Colum. Hum. Rts. L. Rev. 53, 58 (2016). On the normative basis for international responsibility sharing for refugees, see Seth Davis, Responsibility Sharing Within Borders, 110 Calif. L. Rev. 873 (examining the normative basis for international responsibility sharing for refugees). Until the 2015 EU emergency measures, the most significant responsibility-sharing arrangement was the landmark 1979 Comprehensive Plan of Action and Orderly Departure Program in Southeast Asia (CPA). The CPA addressed the influx of refugees from Vietnam, Cambodia, and Laos to neighboring countries following the Vietnam war and the atrocities in Cambodia. It included burden-sharing arrangement as a key component. See, e.g., Schuck (1997), supra; Alexander Betts, Comprehensive Plans of Action: Insights from CIREFCA and the Indochinese CPA 30–31 (UNHCR, Working Paper No. 120, 2006). [].

           [3].     Convention Relating to the Status of Refugees, July 28, 1951, 189 U.N.T.S. 137, 13–14. (entered into force Apr. 22, 1954). Responsibility sharing was not more central to the Convention’s overall scheme due in part to the context in which the Convention concluded. Against the backdrop of the post-World War II trauma and refugee crisis, the main challenge was collaboratively resettling existing war refugees. Underlying the Convention was the assumption that the crisis was temporary and that the international community would never again face a challenge of a similar magnitude that might require new and different collective responsibility-sharing mechanisms. Paul Weis, the first legal advisor for the United Nations High Commissioner for Refugees, wrote in the early 1950s that it was quite unclear whether there would be any need for his agency past 1958. He envisioned the refugee crisis as a highly unusual emergency that would soon be resolved. See Paul Weis, The International Protection of Refugees, 48 Am. J. Int’l L. 193, 220 (1954) (“Throughout the history of international protection of refugees, two antinomies are noticeable: one is the evaluation by international bodies, such as the League of Nations and the United Nations, of the refugee problem as temporary—which has been consistently disproved by historical events. This may well find its explanation in the apparent inconsistency between the hypothesis of a community of peace-loving nations which forms the fundamental premise of these international bodies, and the admission of the permanency of the problem of persons fleeing from persecution by members of this community.”).

           [4].     See Schuck (1997), supra note 2, at 272 (“This burden-sharing norm, however, is manifestly weak. In the international instruments in which it can be discerned, the burden-sharing imperative is essentially precatory and hortatory; even its most energetic scholarly exponents like Goodwin-Gill seem to view it more as a moral aspiration than as a legally binding duty on all states.”).

           [5].     See Michael W. Doyle, Responsibility Sharing: From Principle to Policy, 30 Int’l J. Refugee L. 618, 618 (2018).

           [6].     See Schuck (2013), supra note 2, at 7.

           [7].     Rebecca Dowd & Jane McAdam, International Cooperation and Responsibility-Sharing to Protect Refugees: What, Why and How?, 66 Int’l & Comp. L.Q. 863, 864–65 (2017).

           [8].     See Ilse van Liempt, Maybritt Jill Alpes, Saima Hassan, Sevda Tunaboylu, Orcun Ulusoy & Annelies Zoomers, Evidence-Based Assessment of Migration Deals: The Case of the EU-Turkey Statement 29 (2017), [] (“The EU has long collaborated with countries of origin and transit in the form of migration compacts, readmission agreements and Memoranda of Understanding, but the EU-Turkey Statement is different from prior forms of agreements because of the use of this safe-third-country concept.”). See also the discussion of recent U.S. ACAs in Part III.

           [9].     Article 80 TFEU provides: “The policies of the Union set out in [Chapter 2: Policies on Border Checks, Asylum and Immigration] and their implementation shall be governed by the principle of solidarity and fair sharing of responsibility, including its financial implications, between the Member States. Whenever necessary, the Union acts adopted pursuant to this Chapter shall contain appropriate measures to give effect to this principle.” Consolidated Version of the Treaty on the Functioning of the European Union art. 80, July 6, 2016, 2016 O.J. (C 202) 47.

         [10].     See Part IV and Annex II.

         [11].     See Annex I.

         [12].     Another striking recent example that we will not address in depth is Colombia’s decision to grant protected status to one million undocumented Venezuelan migrants who arrived in the country as a result of the prolonged crisis in Venezuela. See Vanessa Buschschluter, ‘We Gave Venezuelan Migrants a License to Dream’, BBC News (May 17, 2021), [].

         [13].     See generally Ayelet Shachar, Instruments of Evasion: The Global Dispersion of Rights-Restricting Migration Policies, 110 Calif. L. Rev. 967 (documenting, in this contribution to the Symposium, the proliferation of arrangements designed to deflect migrants and avoid triggering legal obligations under international refugee law).

         [14].     The CPA mechanism, the most significant historical example of an operative responsibility-sharing mechanism, covered presumptively recognized refugees. See Betts, supra note 2, at 32.

         [15].     See Suleyman Ali, AU Meet Stresses Need for Solidarity and Responsibility Sharing to Help end Forced Displacement in Africa, UNHCR (Oct. 25, 2019), [].

         [16].     Appl. Stay Pending Appeal at 13, Barr v. E. Bay Sanctuary Covenant, 140 S. Ct. 3 (2019) (No. 19A230).

         [17].     Antony J. Blinken, Press Statement: Suspending and Terminating the Asylum Cooperative Agreements with the Governments of El Salvador, Guatemala, and Honduras, U.S. Dept. State (Feb. 6, 2021), [].

         [18].     See UN High Commissioner for Refugees, Guidance Note on Bilateral and/or Multilateral Transfer Arrangements of Asylum-seekers (May 2013), [].

         [19].     See Application for a Stay Pending Appeal, Barr v. E. Bay Sanctuary Covenant, 140 S. Ct. 1, 3 (2019) (detailing an application for a stay pending appeal of a district court’s injunction against a rule denying asylum to persons who apply in the United States without having applied in the first safe country through which they traveled), []. The application noted that “The [USG] adopted a rule that encourages asylum-seekers to present their claims in the first safe country in which they arrive. That rule . . . is similar to a requirement in effect in the European Union . . . .” Id. at 2–3. It further maintained that “the rule ‘is in keeping with the efforts of other liberal democracies to prevent forum-shopping by directing asylum-seekers to present their claims in the first safe country in which they arrive.’” Id. at 13. But that comparison to the EU ignores the fact that the first safe country requirement in EU migration law is now only one component of a broader scheme that offsets some of its regressive effects (as we will see below). U.S. immigration law lacks comparable balancing arrangements.

         [20].     Implementing Bilateral and Multilateral Asylum Cooperative Agreements Under the Immigration and Nationality Act, 84 Fed. Reg. 63994, 64000 (explaining that “[t]he INA’s ACA provision embodies the policy aim of entering into bilateral or multilateral agreements to promote burden-sharing between the United States and other countries with respect to refugee protection. The U.S. efforts to formulate ACAs with foreign countries is in keeping with the efforts of other liberal democracies to formulate cooperative arrangements in which multiple countries agree to share the review of refugee claims for protection” and citing to UNHCR guidance).

         [21].     Cf. Jaya Ramji-Nogales, Moving Beyond the Refugee Law Paradigm, 111 Am. J. Int’l L. Unbound 8, 8 (2017) (calling for reform that critically evaluates the structure of global migration law more broadly).

         [22].     See, e.g., UNHCR Warns Against “Exporting” Asylum, Calls for Responsibility Sharing for Refugees, Not Burden Shifting, UNHCR (May 19, 2021), [].

         [23].     See, e.g., Michael J. Sandel, What Money Can’t Buy: The Moral Limits of Markets 64 (2012) (“There is something distasteful about a market in refugees, even if it leads to more refugees finding asylum. But what exactly is objectionable about it? It has something to do with the fact that a market in refugees changes our view of who refugees are and how they should be treated. It encourages the participants . . . to think of refugees as burdens to be unloaded or as revenue sources, rather than as human beings in peril.”).

         [24].     See Shachar, supra note 13.

         [25].     UNHCR, Rep. on the Global Compact of Refugees, at 1, A/73/12 (Part II) (Sept. 13, 2018) [hereinafter Global Compact], [].

         [26].     For an overview of the process and citations to related documents, see Volker Türk, The Promise and Potential of the Global Compact on Refugees, 30 Int’l J. Refugee L. 575 (2018).

         [27].     Global Compact, supra note 25, § I.B.5.

         [28].     Id.

         [29].     Global Compact, supra note 25, § III.A.1.

         [30].     Summary of the First Global Refugee Forum by the Co-convenors, UNHCR (Dec. 18, 2019), []. For a subsequent overview of GRF pledges, see Summary of Participation and Pledges at the Global Refugee Forum, UNHCR (Jan. 2020), [].

         [31].     The World Bank contribution includes a component of funding for refugees and host communities over a three-year window ($2.2 billion). An additional $2.5 billion was funding aimed at boosting the private sector and creating jobs for both refugees and host communities. See World Bank Announces $2.2 Billion Scale-up in Support for Refugees and Host Communities at First Global Refugee Forum, World Bank (Dec. 17, 2019), []. The Inter-American Development Bank pledged $1 billion over a period of four years. See Inter-American Development Bank Statement, Global Refugee Forum, UNHCR (Dec. 2019), [].

         [32].     Id.

         [33].     Global Compact, supra note 25, § III.A.2.2.

         [34].     Global Compact, supra note 25, § III.A.3. The Compact outlines rather generic and underspecified tools for supporting the burden and responsibility-sharing mechanisms it envisions. These tools include securing funding and effective resource allocation to affected host countries through state and other actor humanitarian assistance, development cooperation, and maximizing private sector contributions; encouraging partnerships among states, UN actors, other stakeholders, host communities and refugees; and collecting reliable, real-time data. Although the Compact assigns host countries the role of triggering the formation of SPs, the convenors of the first GRF in 2019 announced the launch of SPs to support several regional refugee responses: MIRPS in Central America and Mexico, the Nairobi Process in the East and Horn of Africa, and the Solutions Strategy for Afghan Refugees. These platforms frame their main goal as creating “a mechanism to support responsibility-sharing on forced displacement.” For example, the MIRPS SP coordinates support from MIRPS countries to address the plight of roughly 890,000 people forced to flee from northern Central America and Nicaragua. MIRPS Support Platform, UNHCR (Apr. 22, 2021) []. Its current members are Spain, Argentina, Brazil, Canada, Colombia, the EU, France, Switzerland, Uruguay, and the United States.

         [35].     See, e.g., Lama Mourad & Kelsey P. Norman, The World Is Turning Its Back on Refugees, Atlantic (Dec. 24, 2019), []. See also B. S. Chimni, Global Compact on Refugees: One Step Forward, Two Steps Back, 30 Int’l J. Refugee L. 630, 630 (2018) (“[T]he Refugee Compact is a flawed text for several reasons: it avoids mention of the principal cause of recent refugee flows; dilutes established principles of international refugee law; may weaken the protection of children and women; is short on real mechanisms for responsibility sharing; is myopic in stressing ‘specific deliverables’ . . . in speaking of future academic work; and leaves to the United Nations High Commissioner for Refugees (UNHCR) the task of supervision which it is not equipped to perform (as the Compact itself, which it helped draft, demonstrates).”).

         [36].     See Alexander Betts, The Refugee Compact: Towards a Theory of Change?, 30 Int’l J. Refugee L. 623, 625 (2018). The article surveys historical examples of smaller-scale UNHCR efforts at formalizing responsibility-sharing mechanisms. Id. at 624.

         [37].     Resettlement, UNHCR, (2019) [].

         [38].     Resettlement at a Glance (Jan.-Dec. 2020), UNHCR (Mar. 1, 2021), [].

         [39].     See Key Facts about Refugees in the U.S., Pew Rsch. Ctr. (Oct. 7, 2019), [].

         [40].     See Funding Update 2020, UNHCR, (Dec. 31, 2020) [].

         [41].     See Karen DeYoung, Ruth Eglash & Hazem Balousha, U.S. Ends Aid to United Nations Agency Supporting Palestinian Refugees, Wash. Post (Aug. 21, 2018), [].

         [42].     See Hady Amr, In One Move, Trump Eliminated US Funding for UNRWA and the US Role as Mideast Peacemaker, Brookings (Sep. 7, 2018), []. The Biden administration has restored funding for UNRWA. United States Announces Restoration of U.S. $150 Million to Support Palestine Refugees, UNRWA (Apr. 7, 2021), [].

         [43].     Factsheet: U.S. Refugee Resettlement, Nat’l Immigr. Forum (2020), []. President Trump first reduced the cap on refugee admissions in FY 2017. As a result, the number of admitted refugees declined to fewer than fifty-four thousand. In FY 2018, the president further reduced the refugee admission cap to forty-five thousand, followed by further reductions to thirty thousand for FY 2019, and eighteen thousand for FY 2020. As of September 25, 2020, 10,892 refugees had been resettled in 2020. For Fiscal Year 2021, President Trump issued an Executive Order setting the annual cap at fifteen thousand, the lowest since the enactment of the Refugee Act. See Presidential Determination on Refugee Admissions for Fiscal Year 2020, Memorandum for the Secretary of State, 85 Fed. Reg. 71219 (Nov. 6, 2020). For comparison, in FY 2016, the final year of the Obama administration, the United States admitted nearly eighty-five thousand refugees. See Fact Sheet: Fiscal Year 2016 Refugee Admissions, U.S. Dep’t State (Oct. 4, 2016), [].

         [44].     See Deborah Amos, Biden Plans To Reopen America To Refugees After Trump Slashed Admissions, NPR (Nov. 11, 2020), [].

         [45].     On migration and the Biden administration, see Jaya Ramji-Nogales, Cultivating Normative Authority: The Biden Administration, Migration, and the International Legal Order, 115 Am. J. Int’l L. Unbound 46, 46 (2021); Sarah Libowsky & Krista Oehlke, President Biden’s Immigration Executive Actions: A Recap, Lawfare (Mar. 3, 2021), [].

         [46].     See The Biden Plan for Securing Our Values as a Nation of Immigrants, Biden-Harris, [] (last visited Mar. 21, 2022).

         [47].     See Michael D. Shear, Biden, in Reversal, Raises the Refugee Admission Cap to 62,500 in the Next Six Months, N.Y. Times (May 3, 2021), []. However, the administration fell far short of this number. As of October 2021, the United States only admitted twelve thousand refugees, leaving fifty-one thousand spots unfilled. See Bill Frelick, Biden Administration Falls Far Short of US Refugee Admissions Cap, Hum. Rts. Watch (Oct. 7, 2021), [].

         [48].     Memorandum for the Secretary of State on Presidential Determination on Refugee Admissions for Fiscal Year 2022, White House (Oct. 8, 2021), []; Michael D. Shear, The Biden Administration Will Raise the Cap on Refugee Admissions to 125,000, N.Y. Times (Sept. 20, 2021), [].

         [49].     Those measures include the Migrant Protection Protocol, also known as the “Remain in Mexico” policy, which required asylum seekers who presented at the southern border to remain in Mexico while they awaited an immigration court hearing. The policy was challenged in Innovation L. Lab v. Wolf, 951 F.3d 986 (9th Cir. 2020). Another measure required applicants for asylum in the United States to first apply in the first safe country through which they transited. That policy was challenged in East Bay Sanctuary Covenant v. Barr, 964 F.3d 832 (9th Cir. 2020) and Caital. Area Immigrants’ Rights Coalition v. Trump, 471 F. Supp. 3d 25 (D.D.C. 2020).

         [50].     Fact Sheet: DHS Agreements with Guatemala, Honduras, and El Salvador, U.S. Dept. Homeland Sec. (Aug. 2019), []. The El Salvador and Honduras ACAs were signed in September 2019 but never entered into force. The Guatemala ACA was signed in July 2019 and was implemented later that year. Between November 21, 2019, and March 16, 2020, the United States transferred 939 Honduran and Salvadoran asylum seekers to Guatemala, most of whom were women and children. Only about 2 percent of transferred migrants applied for asylum in Guatemala despite many of them having well-founded fears of persecution in their home countries. See Deportation with a Layover: Failure of Protection Under the US-Guatemala Asylum Cooperative Agreement, Hum. Rts. Watch & Refugees Int’l. (2020), [].

         [51].     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, Nov. 29, 2004, 69 Fed. Reg. 69479, 69479.

         [52].     Matthew Lee, U.S. Restores Some Aid to El Salvador, Honduras, Guatemala, Associated Press (June 18, 2020), []. But see Elisabeth Malkin, Trump Turns U.S. Policy in Central America on Its Head, N.Y. Times (Mar. 30, 2019), [] (stating that only $450 million had been revoked).

         [53].     Id.

         [54].     Press Release, U.S. Dep’t of State, United States Provides Additional U.S. Foreign Assistance for El Salvador, Guatemala, and Honduras (June 24, 2020), [].

         [55].     See Complaint, U.T. v. Barr, No. 1:20-cv-00116-EGS (D.D.C. Jan. 15, 2020); El Salvador Says It Is Not Ready to Receive Asylum Seekers, Al Jazeera (Feb. 6, 2020), (quoting Minister of Foreign Affairs Alexandra Hill Tinoco in February 2020: “We are not going to admit anyone seeking asylum until we as a country have the conditions and technical, financial and human capacity to be able to give these people who are seeking asylum and sent to another country the best treatment.”); Is Honduras Safe for Refugees and Asylum Seekers? Hum. Rts. First 2 (2020), []; U.N. Human Rights Council, Report of the Special Rapporteur on the Human Rights of Internally Displaced Persons on his Mission to Honduras at 1 A/HRC/32/35/Add.4 (Apr. 2016), []; Kevin Sieff & Mary Beth Sheridan, The U.S. Sent Central American Asylum Seekers to Guatemala to Seek Refuge. None Were Granted Asylum, Report Says, Wash. Post (Jan. 16, 2021), []. See also Committee on Implementing Bilateral and Multilateral Asylum Cooperative Agreements under the Immigration and Nationality Act, supra note 20, at 63994.

         [56].     See Barr v. E. Bay Sanctuary Covenant, supra note 19, at 13.

         [57].     See Blinken, supra note 17.

         [58].     ‘After the Airlift’. News Comment Attributable to UN High Commissioner for Refugees Filippo Grandi, UNHCR (Aug. 30, 2021), [].

         [59].     See Afghanistan, UNHCR, [] (last visited Mar. 21, 2022).

         [60].     The United States Conducts Unprecedented Relocation Effort, U.S. Dep’t State (Sept. 6, 2021), [] (stating that the evacuation included “personnel from partner nations, and at-risk Afghan”).

         [61].     Id.

         [62].     Office of Global Partnerships: Partnerships for Afghan Response, U.S. Dep’t State, [] (last visited Mar. 21, 2022).

         [63].     U.S. Refugee Admissions Program Priority 2 Designation for Afghan Nationals, U.S. Dep’t State (Aug. 2, 2021), []. The administration announced a P-2 designation for Afghan nationals. The designation allows Afghans who are ineligible for SIV but worked for U.S. or coalition forces in Afghanistan, those who work for a U.S. government funded program, and those employed in Afghanistan by a U.S.-based media company or non-governmental organization access to USRAP under certain conditions. Id.

         [64].     Hardin Lang, Sarah Miller, Daphne Panayotatos, Yael Schacher & Eric Schwartz, After the Airlift: Protection for Afghan Refugees and Those Who Remain at Risk in Afghanistan 5 (2021), [].

         [65].     Shalanda Young, Delivering for the American People and Meeting Urgent Needs in the New Fiscal Year, The White House (Sept. 7, 2021), [].

         [66].     See, e.g., What’s Next for Afghans Fleeing the Taliban?, Hum. Rts. Watch (Sept. 9, 2021), [].

         [67].     See, e.g., Ellen Mitchell & Rebecca Kheel, Overnight Defense & National Security — Out of Afghanistan, but Stuck in Limbo, Hill (Sept. 3, 2021), []; Nikolas Feith Tan, The Temporary Hosting of Evacuated Afghans in Third Countries: Responsibility Sharing or Externalisation? Refugee L. Initiative (Sept. 15, 2021), [].

         [68].     Cf. Jaya Ramji-Nogales, Silos of International Law: Occupation and Forced Migration, in War, Occupation, and Refugees (Richard Falk & Tom Syring, eds. forthcoming 2021).

         [69].     Cf. Kenneth W. Abbott & Duncan Snidal, Why States Act Through Formal International Organizations, 42 J. Conflict Resol. 3 (1998) (arguing that international organizations gain their power from being able to act as neutral brokers in what would otherwise be political negotiations).

         [70].     See Schuck (1997), supra note 2; Schuck (2013), supra note 2.

         [71].     But see, e.g., Iris Goldner Lang, Is There Solidarity on Asylum and Migration in the EU?, 9 CYELP 1 (2013) (a pre-2015 discussion of limited progress in the implementation of the solidarity principle); Evangelia (Lilian) Tsourdi, Solidarity At Work? The Prevalence of Emergency-Driven Solidarity in the Administrative Governance of the Common European Asylum System, 24 Maastricht J. Eur. & Comp. L. 667 (2017) (arguing that the solidarity principle is insufficiently reflected in the European Asylum system and criticizing the emergency nature of solidarity mechanisms the EU has adopted thus far).

         [72].     See 2015: The Year of Europe’s Refugee Crisis, UNHCR (Dec. 8, 2015), []. See also Melissa Carlson, Laura Jakli & Katerina Linos, Rumors and Refugees: How Government-Created Information Vacuums Undermine Effective Crisis Management, 62 Int’l Studs. Q. 671, 671 (2018); Melissa Carlson, Laura Jakli & Katerina Linos, Refugees Misdirected: How Information, Misinformation, and Rumors Shape Refugees’ Access to Fundamental Rights, 57 Va. J. Int’l L. 539, 566 (2018). A database mapping the journeys of displaced persons across Europe is available at

         [73].     Council Decision 2015/1601, ¶ 17, 2015 O.J. (L 248/80) (EU). The Decision was originally intended to cover Hungary as well, but Hungary declined emergency assistance.

         [74].     Council Decision 2015/1523, ¶ 22, 2015 O.J. (L 239/146) (EU).

         [75].     See Proposal for a Council Decision Establishing Provisional Measures in the Area of International Protection for the Benefit of Italy, Greece and Hungary, at 15, COM (2015) 451 final (Sept. 9, 2015) (“The proposed distribution key should be based on a) the size of the population (40% weighting), b) the total of the GDP (40% weighting), c) the average number of asylum applications per one million inhabitants over the period 2010-2014 (10% weighting, with a 30% cap of the population and GDP effect on the key, to avoid disproportionate effects of that criterion on the overall distribution) and d) the unemployment rate (10% weighting, with a 30% cap of the population and GDP effect on the key, to avoid disproportionate effects of that criterion on the overall distribution).”). See also Joined Cases C 643/15 and C 647/15, Hungary v. Council, ¶¶ 300–01, ECLI:EU:C:2017:631. On the evolution of the EU allocation formulae, see Annex II.

         [76].     Eur. Ct. Auditors, Asylum, Relocation and Return of Migrants: Time to Step Up Action to Address Disparities Between Objectives and Results 23 (2019), [].

         [77].     See Joined Cases C 643/15 and C 647/15, supra note 75.

         [78].     Eur. Ct. Auditors, supra note 76, at 20–21. Of the 40,000 and 120,000 initial relocation target numbers in the 2015 emergency regulations, 7,744 spots out of the 40,000 were never allocated, and 54,000 of the 120,000 were made available for resettling Syrians from Turkey after the 2016 EU-Turkey joint statement, further discussed below. Therefore, Member States were eventually legally required to relocate 98,256 individuals of the original 160,000 target. 34,705 eligible migrants (12,706 from Italy and 21,999 from Greece) were relocated to twenty-two Member States and three associated countries (Liechtenstein, Norway and Switzerland). The United Kingdom and Denmark opted out of the relocation mechanism, exercising their EU Treaty right to do so. Hungary, Poland, and Austria did not relocate anyone. A March 2017 study commissioned by the European Parliament’s Policy Department for Citizens’ Rights and Constitutional Affairs at the request of the LIBE Committee (The LIBE Study) provided data about top and bottom relocators, as well as relative compliance of each member state considering its assigned quota. None of the EU member states met their assigned relocation quota. See Elspeth Guild, Cathryn Costello & Violeta Moreno-Lax, Implementation of the 2015 Council Decisions Establishing Provisional Measures in the Area of International Protection for the Benefit of Italy and of Greece 24 (2017) [hereinafter LIBE Study], []. Malta had the largest compliance rate (61.1 percent), while the other top relocators’ compliance rates ranged between 32.1 and 49.5 percent. The top ten relocators in absolute numbers were France (2,696 relocations), Germany (1,349 relocations), the Netherlands (1,274 relocations), Portugal (922), Finland (919 relocations), Spain (745 relocations), Romania (558 relocations), Ireland (241 relocations), Lithuania (229 relocations), and Luxemburg (226 relocations). Norway (493 relocations) and Switzerland (368 relocations), two Schengen associated States that agreed to take part in the relocation scheme through bilateral agreements, were also among the top relocators. The bottom ten relocators in absolute numbers were Poland, Austria, and Hungary (0 relocations), Slovakia (9 relocations), Liechtenstein (10 relocations), Croatia (19 relocations), Bulgaria (29 relocations), Sweden (39 relocations), Cyprus (65 relocations), and Estonia (78 relocations).

         [79].     Eur. Ct. Auditors, supra note 76, at 2.

         [80].     Germany on Course to Accept One Million Refugees in 2015, Guardian (Dec. 7, 2015), [].

         [81].     See LIBE Study, supra note 78, at 33.

         [82].     See Proposal for a Regulation of the European Parliament and of the Council establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person (recast) at 2, COM (2016) 270 final/2 (May 4, 2016).

         [83].     See Annex I.

         [84].     See Eur. Comm’n, New Pact on Migration and Asylum, Setting Out a Fairer, More European Approach (Sept. 23, 2020), [].

         [85].     See Proposal for a Regulation of the European Parliament and of the Council on Asylum and Migration Management and Amending Council Directive (EC) 2003/109 and the proposed Regulation (EU) XXX/XXX [Asylum and Migration Fund], at 93-94, COM (2020) 610 final (Sept. 23, 2020).

         [86].     Id. at 22.

         [87].     See Michael Doyle, Janine Prantl & Mark Wood, Principles for Responsibility Sharing: Proximity, Culpability and Capability, 110 Calif. L. Rev. 935, 957 (2022).

         [88].     See id.

         [89].     See Shachar, supra note 13. See also Lama Mourad & Kelsey P. Norman, Transforming Refugees into Migrants: Institutional Change and the Politics of International Protection, 26 Eur. J. Int’l Relations 687, 689, 697–98 (2020); Externalizing Migration Management: Europe, North America and the Spread of “Remote Control” Practices (Ruben Zaiotti ed., 2016). Part II of the book is dedicated to externalizing migration management in Europe.

         [90].     On restrictive migration control methods beyond the EU, see, generally, Ayelet Shachar, The Shifting Border: Legal Cartographies of Migration and Mobility (2020).

         [91].     Mourad & Norman, supra note 89, at 697.

         [92].     See European Council Press Release, EU-Turkey Statement, 18 March 2016, [].

         [93].     Id.

         [94].     See, e.g., Van Liempt et al., supra note 8; Amnesty Int’l, A Blueprint for Despair: Human Rights Impact of the EU-Turkey Deal (2017), []; Q&A: Why the EU-Turkey Migration Deal Is No Blueprint, Hum. Rts. Watch (Nov. 14, 2016), []; Elena Becatoros, 3 Years On, What’s Become of the EU-Turkey Migration Deal?, Associated Press (Mar. 20, 2019), [].

         [95].     Susan Fratzke, International Experience Suggests Safe Third-Country Agreement Would Not Solve the U.S.-Mexico Border Crisis, Migration Pol’y Inst. (June 2019), [].

         [96].     See Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions on a New Pact on Migration and Asylum, at 17–18, COM (2020) 609 final (Sept. 23, 2020). See also Communication from the Commission to the European Parliament, the European Council and the Council: Third Report on the Progress Made in the Implementation of the EU-Turkey Statement, at 2, COM (2016) 634 final, (Sept. 28, 2016). The EU has considered similar arrangements with Jordan, Lebanon, Niger, Nigeria, Senegal, Mali, and Ethiopia. See Van Liempt et al., supra note 8, at 29.

         [97].     Joint Press Release: Meeting of the Joint AU-EU-UN Taskforce to Address the Migrant and Refugee Situation in Libya, African Union (Sept. 25, 2019), [].

         [98].     Mourad & Norman, supra note 89, at 697.

         [99].     See 540,000 Migrants Arrive on Greek Islands in the First 10 Months of 2015, ReliefWeb (Nov. 10, 2015), []; van Liempt et al., supra note 8, at 10. In early 2021, Greece sought to send back an additional 1,450 migrants. See Greece Seeks to Send 1,450 Migrants Back to Turkey, Reuters (Jan. 14, 2021), [].

      [100].     See, e.g., John Otis, Colombia’s President on Amnesty for Venezuelans: ‘We Want To Set An Example’, NPR (Mar. 3, 2021), [] (stating that funds for the Venezuelan refugee crisis is relatively less than that for the Syrian or South Sudan refugee crises).

      [101].     See Omer Karasapan, Turkey’s Syrian Refugees–the Welcome Fades, Brookings (Nov. 25, 2019), [].

      [102].     See Kerilyn Schewel, Understanding Immobility: Moving Beyond the Mobility Bias in Migration Studies, 54 Int’l Migration Rev. 328, 338 (2020).

      [103].     See Eur. Ct. Auditors, supra note 76.

      [104].     See Amnesty Int’l, supra note 94, at 6.

      [105].     See UNHCR, supra note 22 (quoting UNHCR’s Assistant High Commissioner for Protection, Gillian Triggs: “I am dismayed by the argument that it is more cost effective to send and host asylum seekers in countries of the global south. I find this morally reprehensible—we must not put price tags on human lives. Refugees are not commodities that can be traded by wealthier nations. To do so is dehumanizing, exploitative and dangerous.”).

      [106].     Cf. E. Tendayi Achiume, Empire, Borders, and Refugee Responsibility Sharing, 110 Calif. L. Rev 1011, 1011 (2022); E. Tendayi Achiume, Migration as Decolonization, 71 Stan. L. Rev. 1509, 1509 (2019).

      [107].     Ukraine Emergency, UNHCR, [].

      [108].     See, e.g., Lara Jakes, For Ukraine’s Refugees, Europe Opens Doors that Were Shut to Others, N.Y. Times (Mar. 3, 2022), [].

      [109].     See Council Implementing Decision 2022/382, Mar. 4, 2022 O.J. (L 71/1) (EU).

      [110].     See Council Directive 2001/55, July 20, 2001 O.J. (L 212/12) (EC).

      [111].     Council Implementing Decision 2022/382, supra note 109, at 3.

      [112].     See Part IV.A.

      [113].     See William N. Evans & Daniel Fitzgerald, The Economic and Social Outcomes of Refugees in the United States: Evidence from the ACS, 47 (Nat’l Bureau of Econ. Rsch., Working Paper No. 23498, 2017), []. See also Steven A. Camarota, The High Cost of Resettling Middle Eastern Refugees, Ctr. Immigr. Studs. (Nov. 4, 2015), [ ] (arguing that it costs far less to resettle Middle Eastern refugees in neighboring Middle Eastern countries than to resettle them in the United States, and estimating that the annual cost per refugee resettled in the United States over the first five years is about $12,900. The study maintains that “in their first five years in the United States each refugee from the Middle East costs taxpayers $64,370—twelve times what the UN estimates it costs to care for one refugee in neighboring Middle Eastern countries”).

      [114].     See Who Bears the Cost of Integrating Refugees?, OECD (Jan. 2017), [].

      [115].     Id. at 2.

      [116].     See also Doyle, Prantl & Wood, supra note 87 (providing data about costs of asylum in the developing world and assessing costs in the hundreds-low thousands per person annually).

      [117].     See Camarota, supra note 113.

      [118].     See Victoria Kelberer, Negotiation Crisis: International Aid and Refugee Policy in Jordan, 4 Middle E. Pol’y 148, 149 (2017).

      [119].     See Otis, supra note 100. See also Dany Bahar & Meghan Dooley, Venezuelan Refugees and Their Receiving Communities Need Funding, Not Sympathy, Brookings (Feb. 26, 2021), [] (“Based on the figures for 2020, total funding per refugee amounts to $3,150 per Syrian, $1,390 per South Sudanese, and just $265 per Venezuelan. In other words, funding for the Syrian refugees has been over 10 times larger than for Venezuelans, in per capita terms.”).

      [120].     Jason Richwine, Steven A. Camarota & Karen Zeigler, The Fiscal Impact of Refugee Resettlement, Ctr. Immigr. Studs. (Mar. 2, 2020), [].

      [121].     See Evans & Fitzgerald, supra note 113.

      [122].     UNHCR, The Impact of Government-Sponsored Refugee Resettlement: A Meta Study of Findings from Six Countries vi (Dec. 2020). See also id. at 32 (surveying U.S. studies showing that resettlement costs are offset by long term earnings).

      [123].     Joakim Ruist, The Fiscal Aspect of the Refugee Crisis, 27 Int’l Tax Pub. Finance 478, 478 (2020).

      [124].     See also Jennifer M. Chacón, Recounting: An Optimistic Account of Migration, 110 Calif. L. Rev. 1041 (2022).

      [125].     See Proposal for a Council Decision Establishing Provisional Measures in the Area of International Protection for the Benefit of Italy, Greece and Hungary, supra note 75. See also Joined Cases C 643/15 and C 647/15, front Council, ¶¶ 300–301, ECLI:EU:C:2017:631.

      [126].     See Council Decision 2015/1601, supra note 73.

      [127].     See Proposal for a Council Decision Establishing Provisional Measures in the Area of International Protection for the Benefit of Italy, Greece and Hungary, supra note 75.

      [128].     See Proposal for a Regulation of the European Parliament and of the Council establishing the criteria and mechanisms for determining the Member State responsible for examining an application for international protection lodged in one of the Member States by a third-country national or a stateless person (recast), supra note 82.

      [129].     Id. at 68.

[130]. Eur. Comm’n, New Pact on Migration and Asylum, Setting Out a Fairer, More European Approach (Sept. 23, 2020), [].


More in this Issue

Empire, Borders, and Refugee Responsibility Sharing

International lawyers have been preoccupied with refugee responsibility sharing for decades, and with good reason. They have invested energy both in critiquing the existing regime and developing proposals for alternatives. However, the corresponding literature has largely, though not entirely, neglected two related but distinct phenomena: imperial domination and imperial intervention. I argue that attention to imperial domination and imperial intervention, which I define shortly, should unsettle the very frame of responsibility sharing and even asylum for many who are coerced into migration.

“Perceived to be Deviant”: Social Norms, Social Change, and New York State’s “Walking While Trans” Ban

Section 240.37 of the New York State Penal Code, colloquially known as the “Walking While Trans” Ban, is an example of our nation’s commitment to its identity—defining the boundaries between what is deviant and non-deviant, what is normative and non-normative. This Note seeks to understand the intersection between criminalization, gender identity, social norms, and the […]

Recounting: An Optimistic Account of Migration

To be forced to move from a beloved home is a tragedy, no matter the cause. But such moves need not end tragically. Though the wounds of losing a homeland may never fully heal, people with the strength and resilience necessary to withstand these kinds of moves are also often well-equipped to build something positive out of pain. They can be tremendous assets to others in their newfound homes.

Responsibility Sharing Within Borders

The international community has long recognized the principle that countries should share responsibility for hosting and supporting refugees. The 2018 Global Compact on Refugees’ recognition of an “urgent need” for greater responsibility sharing across borders reflects widespread agreement that the existing distribution of responsibility among countries is unjust and unworkable. A number of low- and middle-income countries in the Global South are the primary hosts of refugees, while wealthier countries, such as Australia, many European countries, and the United States, deliberately prevent arrivals and strictly limit their acceptance of refugees. The resulting distribution of responsibility tracks the proximity to the countries from which refugees flee, and parallels neither the capacities of states to support refugees, nor the extent of their moral duties to do so.

Principles for Responsibility Sharing: Proximity, Culpability, Moral Accountability, and Capability

Responsibility sharing was a central commitment in the 2016 New York Declaration for Refugees and Migrants.[1] It was also a key commitment in the preamble to the landmark 1951 Refugee Convention, in which countries of first asylum were promised “international cooperation” in return for providing refuge—though the Convention did not specify what this cooperation encompassed. And, just as the 1951 Refugee Convention failed to define what international cooperation meant, the New York Declaration was long on principles but short on specific commitments.

Housing the Decarcerated: Covid-19, Abolition, and the Right to Housing

The coronavirus pandemic revealed the need to advance the right to housing and abolition movements. The need for advancements in both spaces was no more painfully apparent than among the recently decarcerated population. Securing housing for the recently decarcerated is particularly difficult due to the “culture of exclusion” that has long pervaded subsidized housing policy, […]

He Said. She Said. The iPhone Said. The Use of Secret Recordings in Domestic Violence Litigation

This Note explores the use of secret recordings in domestic violence litigation. It is particularly concerned with how the criminalization of domestic violence influences the laws governing the creation and use of secret recordings in this context. Secret recordings can provide determinative evidence of domestic violence. However, a domestic violence survivor who makes a secret recording is criminally and civilly liable under California’s Anti-Eavesdropping Statute (CEPA). CEPA also renders secret recordings inadmissible as evidence.

Instruments of Evasion: The Global Dispersion of Rights-Restricting Migration Policies

This Essay traces the global spread of legal techniques that strive, as official government policy documents explain, to “push the border out” as far away from the actual territorial border as possible. This concept, enthusiastically embraced by governments worldwide, involves screening people “at the source” or origin of their journey, not the destination, and then again at every possible checkpoint along the way. My study is devoted to tracing the core countries and actors that have facilitated the adoption of such policies. I show that desired destination countries systematically learn from, and emulate, each other’s innovations in asylum-denying laws and policies. The lens of diffusion, which guides the analysis, emphasizes processes of inter-jurisdictional learning and emulation. It invites us to ask how and why “ideas travel” across jurisdictions and to trace the complex ways in which states are interacting with one another in shaping their own border control policies.

Protect and Serve

There exists a substantial body of literature on racism and brutality in policing, police reform and abolition, the militarization of the police, and the relationship of the police to the State and its citizenry. Many theories abound with respect to the relationship between the police and Black people in the United States, and most of […]

The Pathological Whiteness of Prosecution

Criminal law scholarship suffers from a Whiteness problem. While scholars appear to be increasingly concerned with the racial disparities within the criminal legal system, the scholarship’s focus tends to be on the marginalized communities and the various discriminatory outcomes they experience as a result of the system. Scholars frequently mention racial bias in the criminal […]