Introduction

In 2015, the world saw the devastating image of the lifeless body of Alan Kurdi, a 2-year-old fleeing the Syrian civil war with his family, washed up on the shores of Turkey, after their inflatable boat capsized trying to cross the Mediterranean to reach a Greek island. Their ultimate destination was Canada because their status in Turkey was precarious, and they had to make the journey irregularly because Canada rejected their request to unite with family members, and Greece had closed its doors.^[1] The current global mobility paradigm, as dominated by the Global North, rests on the aim of stopping and deterring the arrival of asylum seekers and undocumented migrants,^[2] often through preventing departure from countries of origin and transit, which results in countless preventable deaths such as Alan Kurdi’s. Safe third country (“STC”) practices and more broadly, policies for the externalization of migration control are prime manifestations of this containment trend.

Externalization is commonly defined as the strategy of shifting migration control functions, such as border controls or the processing of asylum applications, that are normally undertaken by a state within its own territory, to another state’s territory.^[3] Externalization is problematic because rather than addressing the root causes of migration, it is geared toward its containment, that is, keeping asylum seekers and undocumented migrants as far away as possible. This happens at the expense of their human rights, including the right to seek asylum, the right to leave any country, including one’s own, freedom of movement, and the prohibition of torture, inhuman and degrading treatment.^[4] Some practices of externalization include sanctioning airline companies when passengers land without the required travel documents, outsourcing asylum processing to other countries, and simply sending them to other countries, through first country of asylum or safe third country transfers.^[5]

International refugee law prohibits sending refugees to places where they may face risk of persecution or other serious human rights violations. Still, there is nothing in the Refugee Convention that explicitly prevents sending refugees to safe countries. The safe third country concept is built on this silence and implies that when refugees do not arrive in a state directly from a country where their fundamental rights are at risk, they can be sent back to ‘safe third countries’ they passed through that are willing to accept their return. The STC concept emerged in the late 1980s as a solution to the “asylum shopping” or “refugees in orbit” phenomena, arguably to ensure that refugees do not change countries after they escaped persecution and found protection.^[6] However, in practice, the STC concept tends to cause or aggravate the rights violations asylum seekers face and obstruct their access to asylum. Over the years, the STC concept has been manipulated and stretched so far that states such as Australia and the USA now implement STC transfers to countries that asylum seekers have never even transited through. States use responsibility-sharing as a pretext in relying on the STC concept whereas, in fact, STC practices contain asylum seekers in countries neighboring their country of origin, where they are denied effective protection and are most often incapable of building a future, due to lack of opportunities, discrimination, and undocumented status.

This post approaches STC practices as the embodiment of fundamental problems in the construction of the global asylum regime. According to UNHCR statistics, 67% of refugees and other people in need of international protection live in countries neighbouring the countries they flee from, and 73% of them are hosted in low- and middle-income countries.^[7] STC practices are one of the legal tools that create this unbalanced outcome, compromising refugee protection and global responsibility sharing. As I expand below, the regional impact of the EU-Turkey Statement of 2016 in the aftermath of the dramatic increase in the number of Syrian refugees attempting to reach Europe is a good example of how this plays out in practice. In addition to this concerning distribution of refugees, the STC concept is also problematic because, contrary to the generally accepted view under international law, it assumes that once refugees have escaped the danger in their country of origin, they should seek asylum at the first instance possible and that they do not have the right to choose their country of asylum.^[8] Thus, the STC concept manufactures the ‘illegality’ of secondary movements through legal fiction, created by states. The process surrounding the Canada-USA STC Agreement of 2004 and its 2023 expansion is a manifestation of this dynamic, as shown in the analysis below. These two policy contexts, Europe and North America, have been centers of attraction for scholarly debates around the STC concept since its emergence,^[9] whereas they have remained relatively isolated from each other. Considering they comprise a significant portion of the Global North, an analysis that interlinks them within the asylum research space demonstrates that their STC practices contribute to shaping a common regressive trajectory of the global asylum regime, one that prioritizes the containment of human mobility at the expense of human rights.^[10]

Reinforcement of containment through the STC concept: The EU-Turkey Statement

The STC concept presupposes that asylum should be sought in the first ‘safe’ country that refugees reach directly after leaving their country of origin. This could work if it were in combination with genuine, safe and legal pathways to access asylum in different parts of the world. However, we live in a world dominated by an “architecture of repulsion” aimed at keeping asylum seekers and other migrants away from the Global North.^[11] In view of deterrence measures such as restrictive visa policies and carrier sanctions, it is virtually impossible for refugees to travel directly to distant lands without passing through “safe third countries”. So the STC concept encapsulates a world vision where asylum seekers exclusively remain in or are sent back to the countries neighboring their countries of origin, to which they can travel directly. Combined with other policies of externalisation, this is how the STC practices contribute to a global asylum regime of reinforced containment. This creates tension with the spirit of international cooperation affirmed in the preamble of the Refugee Convention, as the STC concept becomes a tool for ultimate destination countries to avoid responsibility.^[12]

We observe this dynamic in the increased arrival of Syrian refugees triggered by the Syrian civil war after 2011 and the policy interventions shaped by the EU-Turkey Statement of 2016.^[13] After the mass influx started in 2011, Turkey adopted an open border policy and became the top refugee-hosting country in the world with a population of 3.8 million refugees at its peak, triggering the steep increase of often deadly irregular passages through the Mediterranean Sea by 2015.^[14] The policy response to this was the EU-Turkey Statement of 2016, which effectively designated Turkey as a safe third country vis-à-vis the EU. With the Statement, Turkey agreed to prevent irregular migration toward Europe and take back irregular migrants who crossed to the Greek islands from Turkish shores. For each return, the EU committed to resettle one Syrian asylum seeker from Turkey. In addition to the EU’s financial support for refugee protection and containment measures, including detention and border management in Turkey,^[15] another component of the EU-Turkey Statement was the establishment of a Voluntary Humanitarian Admission Scheme, which never materialized.^[16] Almost ten years later today, a total of around 48000 Syrian refugees have been resettled to EU countries from Turkey,^[17] almost 41000 being within the scope of the EU-Turkey Statement.^[18] The dominant portion of Syrian refugees, 4.4 million out of a total of 4.8 million, continue to be hosted in neighboring countries, including Lebanon, Jordan, Iraq, and Egypt, with a population of above 2.5 million in Turkey.^[19] This picture supports the view that the EU-Turkey Statement is a machinery of containment rather than a tool of responsibility sharing.

This leads to a troubling situation. After a few years of open border policy, by 2018, Turkey had built a wall along its border with Syria, which ranks as the 3rd longest wall in the world after the Great Wall of China and the USA-Mexico border wall.^[20] This is a testament to the domino effect created by externalization policies, whereby the tools for externalizing migration control are reproduced by the countries affected by them or ‘partner’ states.^[21] This is also visible in Turkey’s increased efforts to enter into readmission agreements with countries of origin and transit after the adoption of the EU-Turkey Statement. Whereas Turkey executed a total of eight readmission agreements prior to the Statement, five have been executed after, and negotiations are ongoing for two more with Afghanistan and Sudan.^[22]

Manufacturing of ‘illegality’ through the STC concept: Canada-USA Safe Third Country Agreement

At the individual level, the STC concept defies the territorial origins of asylum, stemming from the idea of one’s moral responsibility towards a person who arrives at one’s doorstep and is in need of refuge from danger. The proponents of the STC concept argue that sending refugees to safe countries is legally allowed because it is not explicitly prohibited,^[23] whereas there is no explicit prohibition of the right to choose the country of asylum either. Considering there is no explicit rule in the Refugee Convention, or international law more broadly, that requires individuals to seek protection in the first safe country or any specific country, prioritizing state interests over individual interests is, at the very least, arbitrary. The weight of the authority supports this position, and it is in fact states who are pushing the alternate interpretation and practice.^[24] While the legitimacy of denying asylum seekers agency is highly questionable morally and under international law,^[25] the STC concept acts as a legal tool that creates the so-called illegality of secondary movements despite its promotion as a tool for tackling it. This is how the Safe Third Country Agreement between Canada and the USA (“STCA”)^[26] first created the Roxham Road crossing from the USA to Canada, and now illegalizes its use by asylum seekers.

Until Canada and the USA entered into a Safe Third Country Agreement in 2004, it was possible for asylum seekers who arrived anywhere at the Canadian border to have their asylum claims heard. This allowed asylum seekers to enter Canada in a safe and orderly manner by presenting themselves at an official entry point at the border to seek asylum.^[27] With the STCA and Canada’s designation of the USA as an STC, this changed. Asylum applications submitted at the official ports of entry would no longer be admissible, and applicants could be sent back to the USA. Since the STCA only applied to asylum claims at official border crossings, this caused refugees to seek irregular ways to enter Canada. Hence, the Roxham Road, a country road 50 km south of Montreal and 500 km north of New York City, emerged as the major irregular route from the USA to Canada.^[28] Diversion of what used to be orderly mobility into irregularity was a direct result of the STC arrangement between Canada and the USA. Although the journey for asylum seekers became more dangerous, Roxham Road still provided a relatively orderly passageway for asylum seekers with established reception arrangements. Almost twenty years later, in 2023, the STCA was amended, and the application of the STC concept has now been expanded to cover the entire land border, including unofficial border crossings.^[29] This effectively shut down the Roxham Road for orderly passage of asylum seekers by placing an additional layer of irregularity and thus making their access to asylum more difficult and their journey to safety much riskier. The illegality of mobility is again reproduced through policy intervention.^[30]

First, with the enactment of the STCA, and then with its expansion, we witness a step-by-step thickening and closing of the borders for asylum seekers. The Canada-USA context is the perfect example of the isolated impact of the gradual proliferation of externalization policies through the STC concept. In contrast to the European context, where incoming asylum seekers numbered in the millions, the arrival of fewer than 60,000 between 2017 and 2019 was enough to trigger public perception of a migration ‘crisis’.^[31] There are several factors contributing to this phenomenon, such as spontaneous arrivals being the less usual mode of receiving asylum seekers in Canada.^[32] However, the consequences of policy interventions need to be highlighted. It is, in fact, the STCA itself that diverted refugees into the “unofficial”/“irregular”/“illegal” terrain – creating such ‘illegality’, which in return fuels anti-refugee public sentiment, manufactures a crisis perception, and becomes evidence for further restrictive policies. By pushing asylum claims underground, the STCA also increases the vulnerability of refugees by making their journey more dangerous and opening up a new market for smugglers.^[33] In the end, closing borders does not deter mobility but simply makes it more dangerous.^[34]

Conclusion

This post demonstrated how parallel efforts in the Global North have led to the expanded use of the safe third country concept, showcasing a common pattern in the trajectory of the global asylum regime towards the reinforced containment and illegalization of human mobility. The EU-Turkey Statement highlights the unbalanced distribution of responsibility for refugees among host countries in neighbouring regions through containment policies. On the other hand, developments surrounding the Canada-USA Safe Third Country Agreement reveal a policy architecture that is less conducive to maintaining a protection space for refugees. Looking at the practices of the STC concept, it should be recognized that the problem is not the mobility of refugees trying to seek safety and a better life, but rather the arrangements that contain refugees in conditions where they cannot access effective protection.

Footnotes

1. ‘Family of Syrian Boy Washed up on Beach Were Trying to Reach Canada | Refugees | The Guardian’ <https://www.theguardian.com/world/2015/sep/03/refugee-crisis-syrian-boy-washed-up-on-beach-turkey-trying-to-reach-canada> accessed 25 August 2025. ↑

2. Gamze Ovacık and François Crépeau, ‘Global Compacts and the EU Pact on Asylum and Migration: A Clash Between the Talk and the Walk’ (2025) 14 Laws 3. ↑

3. David Cantor and others, ‘Externalisation, Access to Territorial Asylum, and International Law’ (2022) 34 International Journal of Refugee Law 120. ↑

4. Bachirou Ayouba Tinni and others, ‘Asylum for Containment: EU Arrangements with Niger, Serbia, Tunisia and Turkey’ (ASILE Project 2023) 5; Emilie McDonnell, ‘Challenging Externalisation Through the Lens of the Human Right to Leave’ (2024) 71 Netherlands International Law Review 119, 120–122. ↑

5. Susan Kneebone, ‘The Legal and Ethical Implications of Extraterritorial Processing of Asylum Seekers: The “Safe Third Country” Concept’ in Jane McAdam (ed), Forced Migration, Human Rights and Security (Hart Publishing 2008) 12. ↑

6. ‘Refugee Data Finder - Key Indicators’ <https://www.unhcr.org/refugee-statistics> accessed 26 August 2025. ↑

7. Stephen H Legomsky, ‘Secondary Refugee Movements and the Return of Asylum Seekers to Third Countries: The Meaning of Effective Protection’ 15 International Journal of Refugee Law 566, 570. ↑

8. Nazare Albuquerque Abell, ‘Safe Country Provisions in Canada and in the European Union: A Critical Assessment’ (1997) 31 International Migration Review 569; Michelle Foster, ‘Responsibility Sharing or Shifting? “Safe” Third Countries and International Law’ (2008) 25 Refuge: Canada’s Journal on Refugees 64; Audrey Macklin, ‘Disappearing Refugees; Reflections on the Canada-USA Safe Third Country Agreements’ (2004) 36 Columbia Human Rights Law Review 365; Efrat Arbel, ‘Shifting Borders and the Boundaries of Rights: Examining the Safe Third Country Agreement between Canada and the United States’ (2013) 25 International Journal of Refugee Law 65; Berfin Nur Osso, ‘Unpacking the Safe Third Country Concept in the European Union: B/Orders, Legal Spaces, and Asylum in the Shadow of Externalization’ (2023) XX International Journal of Refugee Law 1; Gamze Ovacık, Meltem Ineli-Ciger and Orçun Ulusoy, ‘Taking Stock of the EU-Turkey Statement in 2024’ (2024) 26 European Journal of Migration and Law 154. ↑

9. Alice Massari, ‘No Countries for Refugees - Canada and Europe’s Shrinking Asylum Space’ (2023) 6 Refugee ‘Responsibility Sharing’ - Challenging the Status Quo: A Special Issue of the PKI Global Justice Journal. ↑

10. David Scott FitzGerald, ‘The Catch-22 of Asylum Policy’ in David Scott FitzGerald, Refuge beyond Reach: How Rich Democracies Repel Asylum Seekers (Oxford University Press 2019) 6. ↑

11. UNHCR, ‘Note on the “Externalization” of International Protection: Policies and Practices Related to the Externalization of International Protection’ para 4 <https://www.refworld.org/policy/legalguidance/unhcr/2021/en/121534> provides that ‘Measures designed, or effectively serving, to avoid responsibility or to shift, rather than share, burdens are contrary to the 1951 Refugee Convention and principles of international cooperation and solidarity. Such externalization measures are distinct from policies and practices aimed at sharing international protection responsibilities in the spirit of international cooperation and solidarity.’ ↑

12. European Council, ‘EU-Turkey Statement of 18 March 2016’ <https://www.consilium.euro pa.eu/en/press/press-releases/2016/03/18/eu-turkey-statement/>. ↑

13. ‘Syria Emergency | UNHCR’ <https://www.unhcr.org/emergencies/syria-emergency> accessed 4 September 2025. ↑

14. ‘Turkey’s EU-Funded Deportation Machine - Lighthouse Reports’ <https://www.lighthousereports.com/investigation/turkeys-eu-funded-deportation-machine/> accessed 15 September 2025. ↑

15. Ovacık, Ineli-Ciger and Ulusoy (n 9) 162. ↑

16. ‘Republic of Türkiye Ministry of Interior Presidency of Migration Management - TEMPORARY PROTECTION’ <https://en.goc.gov.tr/temporary-protection27> accessed 28 August 2025. ↑

17. ‘Eighth Annual Report of the Facility for Refugees in Türkiye’ (European Commission 2024) COM(2024) 593 final 4. ↑

18. ‘Syria Situation Overview - Global Report 2024 | UNHCR’ <https://www.unhcr.org/media/syria-situation-overview-global-report-2024> accessed 4 September 2025. ↑

19. ‘Fact Sheet: The Wall between Turkey and Syria - Global Challenges’ <https://globalchallenges.ch/issue/4/figure/fact-sheet-the-wall-between-turkey-and-syria/> accessed 4 September 2025. ↑

20. Moreno-Lax (n 6) 673–674. ↑

21. ‘Türkiye’de Düzensiz Göç / T.C. Dışişleri Bakanlığı’ <https://www.mfa.gov.tr/turkiye_de-duzensiz-goc.tr.mfa> accessed 28 August 2025. ↑

22. Moreno-Lax (n 6) 667. ↑

23. For a more detailed account, please see Gamze Ovacık, ‘The Right to Choose Country of Asylum: The 1951 Convention and the EU’s Temporary Protection Directive’ in Sergio Carrera and Meltem Ineli-Ciger (eds), EU Responses to the Large-Scale Refugee Displacement from Ukraine: An Analysis on the Temporary Protection Directive and Its Implications for the Future EU Asylum Policy (European University Institute 2023) 126. ↑

24. Frédéric Mégret, ‘The Border Within: Mobility, Stereotypes, and the Case of Asylum Seekers as Migrants’ in Seyla Benhabib and Ayelet Shachar (eds), Lawless Zones, Rightless Subjects: Migration, Asylum, and Shifting Borders (Cambridge University Press 2025) 98–99. ↑

25. ‘Final Text of the Safe Third Country Agreement - Canada.Ca’ <https://www.canada.ca/en/immigration-refugees-citizenship/corporate/mandate/policies-operational-instructions-agreements/agreements/safe-third-country-agreement/final-text.html> accessed 4 September 2025. ↑

26. Audrey Macklin, ‘What Happens When Roxham Road Is Closed’ (27 February 2023) <https://www.thestar.com/opinion/contributors/what-happens-when-roxham-road-is-closed/article_42bbb4d0-54db-5fc8-af4b-50533bd9786f.html>. ↑

27. Karine Côté-Boucher, Luna Vives and Louis-Philippe Jannard, ‘Chronicle of a “Crisis” Foretold: Asylum Seekers and the Case of Roxham Road on the Canada-US Border’ (2023) 41 Environment and Planning C: Politics and Space 408, 410. ↑

28. ‘Additional Protocol to the Safe Third Country Agreement - Canada.Ca’ <https://www.canada.ca/en/immigration-refugees-citizenship/corporate/mandate/policies-operational-instructions-agreements/agreements/safe-third-country-agreement/additional-protocol.html> accessed 5 September 2025. ↑

29. Please see Kate Motluk’s post “Is Trump’s America Safe for Asylum-Seekers? Re-Examining the Safe Third Country Agreement Between Canada and the USA in 2025” in this series for a more detailed analysis including the litigation process before the Supreme Court of Canada. ↑

30. Côté-Boucher, Vives and Jannard (n 28) 409. ↑

31. Audrey Macklin and Joshua Blum, ‘Country Fiche: Canada’ (2021) 22. ↑

32. Massari (n 10). ↑

33. Christina Clark-Kazak, ‘“Responsibility-Sharing” and the Safe Third Country Agreement: From a Politics-Based to a Rights-Based Approach to Refugee Protection’ (2023) 6 Refugee ‘Responsibility Sharing’ - Challenging the Status Quo: A Special Issue of the PKI Global Justice Journal. ↑