Jump directly to page content

The Externalisation of European Refugee Protection

A Legal, Practical and Political Assessment of Current Proposals

SWP Comment 2024/C 13, 28.03.2024, 8 Pages


Research Areas

In many European countries, the number of asylum applications continues to rise, as does the appeal of right-wing populist parties. In response, initiatives such as the Brit­ish government’s Rwanda Plan and the Italian government’s agreement with Albania aim to significantly reduce the number of arrivals by transferring asylum procedures and refugee protection to third countries. It is worth noting that although similar pro­posals in the past have never progressed beyond the idea stage, concrete implemen­tation procedures are currently under discussion for Rwanda and Albania. However, there are several legal and normative concerns as well as practical challenges that need to be carefully considered. These approaches would fundamentally jeopardise international refugee protection and harm vital foreign policy interests as well as the credibility of the development cooperation of Germany and the European Union (EU).

In 2001, the Australian government initi­ated the “Pacific Solution”, which involved the transfer of individuals arriving by boat to camps in Papua New Guinea and Nauru to prevent them from claiming asylum on Australian territory. Although this policy has been effective in preventing irregular crossings to Australia, it has been criticised for resulting in serious human rights abuses. Despite this, in 2003 the United Kingdom (UK) proposed transferring asylum seekers to “Regional Protection Areas”. A year later, former German Interior Minister Otto Schily proposed the establishment of simi­lar centres in North Africa.

In 2018, the European Council also dis­cussed the related concept of “disembarkation platforms”. However, these plans did not materialise due to fundamental doubts about their feasibility in Europe and their incompatibility with international norms.

Currently, there is a renewed interest in the externalisation of asylum procedures and protection to third countries. In Ger­many the Federal Government is currently exploring the feasibility of determining protection status abroad in accordance with the 1951 Refugee Convention and the Euro­pean Convention on Human Rights (ECHR). In their draft programme, the Christian Democratic Union (CDU), which is the main opposition party, has proposed that asylum applications be accepted only in third coun­tries, citing the progress made by the UK and Italy in reaching such agreements with Rwanda and Albania. This claim has also made it into the political manifesto of the centre-right European People’s Party – the pan-European party of the CDU and the Christian Social Union (CSU) – and has sparked divisions within the European Com­mission ahead of European Parliament elections.

The current debate

In December 2023, the EU member states, the European Parliament and the Commission agreed on a “New Pact on Migration and Asylum”. Among other things, new border procedures should lead to a reduc­tion in asylum applications and a significant increase in returns. Additionally, it proposes a complex system of flexible soli­darity to improve cooperation between member states. However, it remains ques­tionable whether and when a reduction in irregular border crossings can actually be achieved while also respecting the funda­mental right to asylum. The new legis­lation, which will be implemented from 2026 onwards at the earliest, may face chal­lenges in achieving this. Given the ongoing crises in the main countries of origin, it is anticipated that the number of asylum appli­cations in the EU will remain high. With regard to the ongoing conflict in Ukraine, the prospect of Ukrainian refugees returning to their home country is remote.

In this context, the externalisation of asylum procedures has three main objec­tives. Firstly, it is aimed at discouraging those who are considering unregulated migra­tion or seeking protection from under­taking the dangerous journey to request asylum in the EU. Secondly, it should facilitate repatriation in the event the asylum request is denied. Thirdly, it is intended to signal that decisive action is being taken against unregulated migration and the smuggling organisations involved.

Furthermore, proponents of externalisation are advancing human rights arguments: The expected reduction in migration would lead to a decrease in deaths on dangerous routes, particularly in the Mediterranean. In the medium term, a new EU system for refugee protection would transition away from individual asylum applications at the external borders towards quotas for those in need of protection who would be accepted directly from third countries. Advocates argue that this approach could benefit vul­nerable people, who are less likely to make the dangerous journey to Europe.

This approach appears to offer a potential solution to the systemic crisis in Euro­pean asylum policy. However, there is no conclusive evidence for the alleged deter­rent effect of outsourcing asylum procedures, and significant legal, practical and political objections remain. It is also far from certain that, in the event of a reduc­tion in irregular migration, EU member states would be willing to accept sufficiently large quotas to accommodate those in need of protection. This would just as likely lead to a marked decrease in the overall number of people granted protection and amount to further restrictions on the right to asylum.

Comparing externalisation approaches

The various efforts to externalise responsibility for people in need of protection share two elements: Firstly, they involve individ­uals who have already arrived in the terri­tory of a state where they want to claim asy­lum, or in the case of sea rescue, are under the effective control of the respective state actors. Secondly, these individuals are to be transferred to a third country in an organ­ised manner. Although there are notable differences what specific tasks governments outsource, three distinct types of externalisation approaches are emerging in practice.

Type 1: Extraterritorial asylum procedures

This type involves the physical relocation of asylum procedures to third countries while still applying the law of the externalising state. The best-known historical example is the Pacific Solution, which Australia im­plemented from 2001 to 2007 in Nauru and Papua New Guinea, despite massive human rights violations. If refugees received a posi­tive decision on asylum, they were to be brought to Australia, although some were transferred to third countries. The recent agreement between Italy and Albania simi­larly foresees the externalisation of asylum procedures, with Italian law being applied throughout. In the event of a positive deci­sion on asylum, protection should only be granted in Italy.

Type 2: Transfer of responsibility for procedures and protection

A second type involves the legal transfer of asylum procedures to third countries, in addition to territorial transfer. Aside from the Australian practice of “offshore pro­cessing”, the most prominent current exam­ple is the agreement between the UK and Rwanda. Under this agreement, asylum seekers transferred from the UK to Rwanda would be subject to Rwanda’s asylum law and receive protection there if their appli­cations are approved. This places considerable demands on Rwanda’s asylum system.

The asylum cooperation agreements between the United States (US) and El Salvador, Guatemala and Honduras, which were in effect from 2019 to 2021, were designed similarly: They aimed to transfer the responsibility for conducting asylum procedures and granting protection in the event of a positive decision entirely to the partner state. However, the first two agree­ments were never implemented.

Type 3: Return to transit countries

A third type of agreement includes pro­visions for the return of those seeking pro­tection to the transit countries through which they have passed. The most well-known example is the EU-Turkey Statement of 2016. Under these types of agreements, it is assumed that adequate conditions for pro­tection exist in the transit state, but there are no specifications about how the asylum procedure will be organised or what basic rights should be guaranteed. The transit state’s consent to repatriate certain groups of people is typically acquired using various incentives. For example, the EU-Turkey State­ment in theory obligated European member states to accept one recognised Syrian refugee for every Syrian returned to Turkey, although this scenario never ma­teri­al­ised. The arrangement also included the prospect of lifting visa requirements for Turkish citizens. However, the EU’s sub­stan­tial financial support given for the re­ception and integration of refugees in Tur­key ultimately proved to be the deciding fac­tor. As part of the European asylum reform, the EU is likely to try to persuade other tran­sit countries to cooperate in a similar way.

International law and human rights obstacles

Delegating state responsibilities for asylum procedures and the granting of protection to third countries is difficult to reconcile with international law. Signatories to the 1951 Refugee Convention are obliged to contribute to the protection of refugees in a spirit of international cooperation and solidarity. Although signatory states do not have to provide territorial asylum to all those seeking protection, the Refugee Con­vention does require them to maintain certain standards regarding all protection options, such as resettlement, safe third-country arrangements and similar mechanisms. This responsibility applies not only to the principle of non-refoulement (Article 33), but also to the effective provision of social and economic rights. It is important to note that Article 3 of the United Nations Convention Against Torture also prohibits the expulsion, extradition or return of indi­viduals to a country where they may be sub­jected to torture, inhuman treatment or serious human rights violations.

Additionally, the ECHR guarantees the right to protection against refoulement. In the case of vulnerable persons, especially minors, special care and duties of protection must be applied. It is worth noting that the ECHR remains binding even if an indivi­dual is located outside of Europe. All con­tracting states must adhere to their human rights obligations as soon as they exercise “effective control” over an individual. This obviously applies to asylum seekers who have already reached the territory of an ECHR state, but also to individuals rescued at sea by the authorities of an ECHR signa­tory state.


Externalisation approaches


Involved countries (destination country & third country)

State of implemen­tation
(26 Feb. 2024)

Place / jurisdiction of asylum determination

Responsibility / location of protection

Responsibility / location for rejected asylum claims

“Pacific Solution” / Type 1

Australia & Nauru / Papua New Guinea (PNG)


PNG & Nauru / Australian jurisdiction

Australia / Resettle­ment in third countries (e.g. New Zealand, Sweden, Canada)

Australia / mostly lasting detention in PNG & Nauru

Italy-Albania proto­col on migration management /
Type 1

Italy & Albania

planned duration 2024–2029

Albania / Italian jurisdiction

Italy (transfer of accepted asylum seekers)


“Offshore pro­cessing” / Type 2

Australia & Nauru / PNG

Since 2012, on pause

PNG & Nauru /
local jurisdiction

Resettlement in third countries (USA, Cambodia)

Australia / mostly lasting detention in PNG & Nauru

US asylum co­operation agree­ments / Type 2

USA / Guatemala & Honduras & El Salvador

no implementation in Honduras and El Salvador

Other respective third countries, Guatemala for citi­zens of Honduras and El Salvador

Other respective third countries, Guatemala for citizens of Hon­duras and El Salvador

Other respective third countries, Guatemala for citizens of Hon­duras and El Salvador

UK-Rwanda Agree­ment for the Pro­vision of an Asylum Partnership & Danish Law L 226 / Type 2

UK (and Den­mark) & Rwanda

Ongoing legislative process, planned duration until April 2027

Rwanda / Rwandan jurisdiction



“EU-Turkey State­ment” / Type 3

EU & Turkey

Since 2016, partially implemented

Turkey / Turkish juris­diction for Syrians, no proper asylum procedures for other national­ities

Turkey, case by case resettlement of particularly vulnerable per­sons to the EU

Turkey / deportations to third countries (Syria, Iran)

The political debate in the UK and the decisions of the UK Supreme Court demon­strate that the process of transferring asylum procedures to Rwanda faces severe legal challenges under the ECHR. As a result, the UK government has instructed civil servants and national courts to declare Rwanda a safe country and to block appli­cations to the European Court of Human Rights. This approach raises alarms about undue interference with the separation of powers. Leaders of the Conservative Party have for some time now been calling for a withdrawal of the UK from the ECHR. This would seriously damage the human rights regime in Europe, which is already under severe stress.

Even beyond fundamental human rights concerns, there are major legal gaps in the current externalisation plans. For example, there are inadequate regulations on what happens to people whose asylum applications are rejected. If they can neither return nor be naturalised locally, they – and par­ticularly their children – risk becoming undocumented, or even stateless. The agree­ment between the UK and Rwanda states that “relocated persons” should be granted freedom of movement and personal docu­mentation. However, in practice, even refu­gee documents do not fulfil the function of national passports and cannot ensure regular international mobility.

Practical hurdles

The operationalisation of externalisation plans is extremely challenging. In the case of Denmark, the proposed outsourcing of asylum procedures to Rwanda has not yet been implemented, despite the adoption of a law to that effect two years ago. The main obstacles appear to be related to identifying suitable partners, managing costs and en­suring the scalability of the approach.

Challenging search for partners

While many industrialised destination coun­tries have expressed interest in out­sourcing their protection responsibilities, there is a noticeable lack of third countries willing to collaborate. The Australian gov­ernment has been successful in persuading the politically and economically weak states of Nauru and Papua New Guinea to cooper­ate, as well as Cambodia to a limited extent in the area of resettlement. However, the situation in the European neighbourhood and in Africa is not comparable. Except for the agreements made with Rwanda and Albania, European states have not made any progress on similar projects. For in­stance, the 55 member states of the African Union have expressed their opposition to extraterritorial detention centres on their continent. This is partly due to the lack of public support in potential partner coun­tries.

Autocratic regimes may be less affected by the pressures of public opinion. Yet, they tend to decisively pursue their inter­ests vis-à-vis European governments. It is important to consider that aid money alone may not be sufficient to encourage exten­sive cooperation on migration and asylum policy in these countries. The case of the EU-Turkey Statement shows that strong political incentives and the partner state’s own interests must be aligned. In 2016, Turkey was already handling a significant number of Syrian refugees, and European support was appreciated at the time. The agreements that have recently been struck with Tunisia and Egypt, by contrast, high­light how much political and financial capital the EU is willing to invest in order to guarantee cooperation with stemming the flow of irregular arrivals in Europe. Taking back rejected asylum seekers from third countries – or even allowing asylum procedures to be transferred to their terri­tories – is not something these countries would currently agree to. At most, the part­ner country may agree to expedite the readmissions of its own citizens.

Even when it is possible to agree on the externalisation of asylum procedures, part­ner states may still have concerns. For example, the UK’s agreement with Rwanda permits the transfer of arriving asylum seekers from the UK. Yet, approval must be granted by the Rwandan authorities for each individual person.

High cost and low scalability

The cost-benefit calculation of externalisation plans is highly skewed. The UK gov­ern­ment’s expenditure on its Rwanda agree­ment – so far amounting to £240 million, with an anticipated additional £150 million by 2027 – stands in stark contrast to the absence of any asylum seeker transfers to date. The projected cap of 1,000 transfers by 2027 further underscores the inefficiency of such investments.

Significant additional costs will be in­curred during implementation. For exam­ple, the UK model provides for irregularly arriving migrants seeking protection to be flown out to Rwanda – at high costs. Because of the administrative procedures involved, these individuals are likely to spend several weeks in the UK receiving accommodation and support. The justification for these medium-term expenditures hinges on the potential of having a signifi­cant deterrent effect on immigration flows in the future. Yet, the disparity in spending is still evident in Australia’s commitment of A$485 million in 2023 to the operation of detention centres on Nauru for a mere 22 individuals. This is a stark imbalance in terms of costs and benefits, especially when com­pared to the positive impact these funds could have if spent in the main refugee hosting countries.

In any case, the scalability of outsourcing asylum processing in the European context remains questionable. Europe is not an isolated location with exclusively maritime borders. As the EU handled more than a mil­lion asylum applications in 2023, the feasibility of externalising the processing of all applicants (or even a significant subset) that would be necessary to achieve the pro­posed deterrent effect appears unrealistic. Europe’s geographical and geopolitical realities – bearing closer resemblance to those of the US than to Australia’s – sug­gest limited success in achieving substantial and sustained reductions in arrivals through a further tightening of border controls and efforts to outsource asylum procedures.

Distinct challenges of the agreement with Albania

Could intra-European externalisation along the lines of the Italy-Albania agreement solve some of these problems? There are some reasons to think so. The Italian and Albanian parliaments have now approved the bi­lateral agreement, and the Albanian Consti­tutional Court and the outgoing European Com­mission have also given the green light. Both countries are parties to the ECHR, ensuring fundamental rights. Furthermore, the agreement stipulates that Italy will remain fully responsible for conducting asy­lum procedures for individuals transported to Albania, and – by implication – must adhere to EU standards in secondary law for remedies, hearings and legal assistance throughout.

Nevertheless, the model’s practical im­plementation raises a number of unsolved questions. To begin with, the feasibility of conducting vulnerability assessments on ships, particularly of individuals to be directly transferred to Albania, is highly questionable. The practice of detaining asylum seekers, either during their appli­cation process or following a rejection, ventures into even murkier legal territory. Although not explicitly prohibited, the EU regards such detentions as measures of last resort, echoing the concerns of numerous NGOs about the potential detriment to the mental health and fundamental human rights of the individuals involved. Even so, the Italian model provides for asylum seekers in Albania to be detained for the duration of the procedure. This becomes all the more problematic the longer the pro­ceedings take.

Complications further arise concerning the repatriation of individuals whose asylum requests have been denied. The responsi­bilities and procedures for enforcing such returns, especially in non-voluntary cases, and addressing the impediments to repat­riation – such as lack of documentation, health issues or non-cooperation from ori­gin countries – remain undefined. Given that EU-wide, voluntary or enforced return rates of rejected asylum seekers remain low, similar challenges are to be anticipated in third-country contexts. There­fore, it is likely that asylum seekers transferred to Albania would endure ex­tended periods under detention-like conditions, with a significant possibility of eventually being readmitted to Italy, as stipulated in the bilateral agree­ment. Moreover, it is conceivable that rejected applicants would leave Albanian facilities without authorisation and in­advertently fuel irregular secondary migra­tion within Europe.

A critical evaluation of the economic rationale and strategic validity of the Italy-Albania agreement further underscores its limitations. The Italian government has allocated approximately €650 million for a five-year initiative that is capable of accom­modating 3,000 male asylum seekers. Assuming an ideal scenario in which all procedures – including potential repatriations or transfers to Italy – are concluded within a single month, up to 36,000 appli­cations annually could theoretically be processed in Albania. Yet, even under this optimistic scenario, less than one-fifth of the current irregular arrivals to Italy via the Mediterranean would be transferred to Albania. Additionally, migratory flows could be redirected to other routes, poten­tially burdening other EU member states, thereby questioning the model’s exemplari­ness and broader applicability.

The implications of expanding the externalisation model to other EU accession candidates

In the hypothetical case that the Albanian externalisation model is expanded to other EU accession candidates, this could lead to a fragmented system and severe coordination challenges. A scenario in which countries establish individual agreements to out­source asylum procedures – for instance Germany with North Macedonia, France with Moldova or the Netherlands with Geor­gia – would create serious problems. One key issue would be the emergence of diver­gent standards, as these bilateral agreements might follow the example of the Albania agreement and prioritise national laws over the unified framework of EU legislation. This diversity in legal standards could poten­tially undermine the integrity of the Common European Asylum System (CEAS).

Another option would be the establish­ment of a standardised EU framework agree­ment for externalising asylum pro­cedures. However, this would require mem­ber states to agree on a division of responsibilities and sharing the immense costs. In light of recent experiences in the multi-year negotiations on reforming the CEAS, this seems highly unlikely. And given the cur­rent level of European integration, the direct administration of individual asylum applications by the EU – potentially man­aged by the Agency for Asylum – appears neither feasible nor legally possible. Finally, EU accession candidates could leverage new externalisation agreements to negotiate political concessions, potentially complicating the EU’s strategic objectives for orderly enlargement and adherence to rule-of-law principles.

Conclusion: Reassessing the efficacy and legality of externalisation strategies

The anticipated benefits of externalisation strategies, such as mitigating the loss of lives at EU borders and dismantling smuggling networks, remain largely unproven. Aside from Australia’s unique context, there is scant evidence supporting the success of such policies. Although the EU’s agreement with Turkey has contributed to some reduc­tion in irregular migration via this route, the influx of Syrians seeking refuge in the EU continues at a significant rate, accompanied by an increase in the number of illegal pushbacks.

EU law and the ECHR also impose constraints on overly restrictive policies. Many proposed externalisation initiatives risk violating existing legal frameworks and setting unrealistic political expectations. Nonetheless, proponents of externalisation may seek to navigate legal ambiguities, as seen in the agreement between Italy and Albania. Such approaches often weaken pro­tection and human rights standards, undermining the fundamental right to asylum.

Across the EU, the discourse is currently dominated by proposals for the externalisa­tion of asylum procedures (“Type 1”) and/or protection (“Type 2”). In practice though, “Type 3” agreements, which are focused on coordinated returns to transit countries, remain the most feasible, but they are con­tingent on cooperation from those coun­tries. There is a clear risk that bilateral agree­ments with third countries show a lack of faith in the CEAS and can undermine the functionality of reforms of the CEAS before they are implemented.

More broadly, these externalisation plans threaten to exacerbate the imbalance in the sharing of responsibility between EU mem­ber states and the third countries – most of which are part of the so-called Global South – expected to host and protect refugees. This could have severe repercussions, since these lower-income countries might adopt similar tactics, further under­mining the objectives the Global Compact on Refugees.

Instead of pursuing externalisation approaches and putting even more faith in the cooperation with third countries, the EU should concentrate on the implementation of the CEAS reform and defend the requirement to respect protection standards in the EU. Migration partnerships with third countries need to prioritise strengthen­ing capacities in hosting states and expand­ing safe, regular pathways for those seeking protection as well as labour migrants, stu­dents and trainees. Furthermore, facilitating regional mobility in the so-called Global South could contribute towards creating prospects for affected people in their regions of origin. The current restrictive policies only threaten to deflect attention from the difficult, but necessary work of finding legitimate and sustainable solutions for the real challenges posed by migration and forced displacement.

Dr Steffen Angenendt is a Senior Fellow of the Global Issues Research Division at SWP. Nadine Biehler, David Kipp and Dr Anne Koch are Associates in the Global Issues Research Division. Dr Raphael Bossong is Deputy Head of the EU/Europe Research Division.
This article was written as part of the independent research project Strategic Refugee and Migration Policy, funded by the German Federal Ministry for Economic Cooperation and Development.

© Stiftung Wissenschaft und Politik, 2024


Stiftung Wissenschaft und Politik

ISSN (Print) 1861-1761

ISSN (Online) 2747-5107

(English version of SWP‑Aktuell 12/2024)