The 'safe third country' concept turns out to be an empty shell

On 11 July 2026, research circulated by the Phys.org wire drew attention to a finding that European policymakers would prefer to leave unstated: the legal device known as "safe third country" has, in its only large-scale operational test, rested on a category mistake. The researcher, Dr. Gaia Romeo, walks the reader through the mechanics of returns under the doctrine and finds that the very states being designated "safe" are not, in the operational sense the law requires, ready to absorb the people being sent back to them.

That conclusion matters because the doctrine is the load-bearing pillar of the European Union's recent push to externalise asylum processing. If "safe third country" cannot be made to mean what its proponents claim it means, the policy scaffolding erected on top of it, bilateral readmission deals, accelerated border procedures, scaled-up detention capacity at the Union's external edge, is decorative. The political argument that asylum claims can be disposed of cheaply and quickly by transferring them to a willing third state requires that willing third states actually exist. Romeo's research suggests they do not.

How the doctrine is supposed to work

The legal premise is straightforward on paper. Under the EU's recast Asylum Procedures Directive and the case law of the Court of Justice of the European Union, a member state may reject an asylum claim without examining its substance if the applicant can be returned to a country where they are not at risk of persecution and where they can, in turn, receive effective protection. The receiving state is "safe" by definition, and the applicant is expected either to lodge a claim there or to be re-admitted into a status they already hold.

In practice, three things have to be true for the system to function. The designated country has to have a functioning asylum procedure of its own. It has to consent, in some form, to taking the person back. And the person themselves has to have a real connection to the country, so that returning them is not, in effect, dumping them into stateless limbo. Each of these conditions is harder to satisfy than the policy literature suggests. The state of the academic literature, as compiled in the European Union Agency for Asylum's 2024 country-of-origin briefs, is that very few candidate third countries tick all three boxes simultaneously.

The case that actually got tested

Romeo's study homes in on the single instance in which the doctrine has been applied at scale: returns from European territory to a neighbouring transit state that hosts a large displaced population of its own. Without naming the country in the headline framing, the research describes a jurisdiction whose asylum system is partly suspended, whose domestic absorption capacity is saturated, and whose government has signed readmission understandings in Brussels while issuing public statements in its own capital that effectively disclaim responsibility for the welfare of returnees once they cross back in.

The mismatch is not subtle. The European Commission has signed memoranda of understanding with the government in question, and the Council has cited those memoranda as evidence that the partnership is operational. On the ground, the international organisations that monitor reception conditions describe a system that is at best patchy and, for several categories of returnees, non-functional. Romeo's contribution is to lay out, step by step, where the legal characterisation and the operational reality diverge, and to do so with reference to publicly available monitoring reports, court rulings, and the bilateral instruments themselves.

The political economy of an empty category

There is a temptation to read this as an academic finding with no political consequences. That reading is wrong, and worth saying plainly. The safe third country concept is the technical-legal fig leaf under a much larger political bargain: that European electorates can be told asylum decisions are happening somewhere, even if the somewhere in question is not strictly within the Union. If the fig leaf is taken away, the bargain becomes harder to defend in its current form. Governments that have invested political capital in the externalisation model, including in Berlin, Vienna, The Hague, and Rome, will be obliged either to revise the model or to argue, in public, that operational dysfunction is an acceptable price for reduced arrival numbers.

There is a counter-argument, and it should be taken seriously. Returns to the designated country in question have continued under diplomatic pressure, and the volume of new arrivals on the relevant European border route has fallen. The political objective of the policy, deterrence, in the language of interior ministries, is being met in the narrow sense that fewer people are reaching European territory to claim asylum. Whether that is the same thing as a humane and lawful system is a different question, and it is the one Romeo's research is asking.

What remains uncertain

The honest caveat is that the evidence base on returns under the doctrine remains thin. Independent monitors have limited access to the transit-state facilities in question, and the figures released by European agencies on return outcomes are not, in the researcher's framing, broken down in a way that allows external verification of what happens to returnees after the European side of the process ends. A finding that the doctrine is operationally empty in one major case is a serious indictment. It is not, on the available evidence, a finding that applies uniformly to every potential designation, and the policy debate will turn in part on whether other candidate third countries are better positioned to deliver what the law demands.

The most useful next step, on the reading this publication finds most persuasive, is for the European Parliament's civil liberties committee and the European Court of Auditors to publish their own assessments of return outcomes. If the operational gap Romeo identifies is real and structural, it will show up in those audits within a year. If it is narrower than the research suggests, that evidence will show up too. Either way, the category of "safe third country" is no longer something the European policy debate can take on faith.

This piece is filed under Monexus's science desk but reads as policy: the research record is being used, by politicians in several member states, to make legal claims that the data on the ground does not support. We have led with the researcher's findings and given the deterrence argument its strongest framing, because the honest answer to "does the doctrine work?" should be the one the evidence supports, not the one any government wants to hear.