The 'safe third country' idea is starting to crack under its own weight

On a stretch of southern European coastline that has handled more than a decade of arrivals, a recently completed research project has stripped the policy down to its working parts. Dr. Gaia Romeo's study of the only large-scale instance in which the European Union's "safe third country" concept has actually been applied exposes a legal apparatus that, in her telling, was constructed for speed and now governs human movement on the strength of its own categories.

The finding lands at a moment when the concept is being quietly embedded into the next iteration of the Common European Asylum System. A doctrine originally designed to transfer asylum seekers to a country presumed safe is, on the available evidence, a procedural convenience with little empirical scaffolding behind it. That is the story this publication follows.

The case that exposes the gap

Romeo's research centres on the only case in which the "safe third country" policy has been deployed at scale. The details of jurisdiction, year and recipient country matter less than the structural finding: a category meant to short-circuit lengthy status determinations has, in operation, become the primary tool by which applicants are funnelled out of formal procedure and into administrative return. The academic's argument is that the label does no real safety work; it does relocation work.

That distinction matters in Brussels, where the safe-country concept is being extended, not retired. The next iteration of the Common European Asylum System treats presumption-of-safety lists as an efficiency lever. Romeo's evidence suggests that efficiency has come at the cost of any meaningful assessment of what happens to a transferred person once the file is closed.

What the policy actually does

Read literally, a "safe third country" designation is meant to allow an EU member state to declare that an asylum seeker could have found protection in a country they transited, and to send them back there without examining their substantive claim. The instrument presupposes a chain of functioning asylum systems in transit states, plus diplomatic readmission arrangements that allow removals to actually occur.

Romeo's case study traces how that chain works in practice. The finding, drawn from the single large deployment examined, is that the concept operates as an administrative bypass. Officials apply the label, the applicant's file is closed without substantive consideration, and the actual return depends on bilateral logistics that exist separately from the designation itself.

Two consequences follow. First, decisions on safety are made by the sending state, not by an independent evaluator. Second, when the receiving country cannot or will not readmit, the applicant ends up in a procedural limbo that the designation was supposed to prevent.

The structural mismatch

The deeper problem is not corruption or incompetence; it is design. The "safe third country" instrument was built to handle a smaller, more orderly flow of mixed-movement migrants than the one Europe has actually received in the past decade. Asylum systems in transit states are uneven, and readmission agreements are politically contingent. A presumption-of-safety list assumes away both.

Brussels has been patching this for years. The 2024 Asylum Procedure Regulation tightened the criteria for designations; the 2025 Crisis Regulation added operational flex for member states under pressure. Neither addressed the core empirical question: which transit countries actually deliver protection at a level that would justify sending a person back there?

Romeo's research does not answer that question across the whole map of transit states. It answers it for the one case where the policy has been used at scale, and the answer is uncomfortable. Safety was presumed; whether it was present, in the sense a litigant could enforce, was a separate matter.

What Brussels does next

The pressure on the concept will not let up. Migration is a frontline political issue across most EU member states, and the Commission's incentive is to provide instruments that allow governments to demonstrate administrative throughput. The "safe third country" label is administratively attractive precisely because it converts a substantive question ("is this person in need of protection?") into a procedural one ("could they have got it elsewhere?").

The risk, on this publication's reading, is that the redesign entrenches the instrument rather than re-examines it. If the next Common European Asylum System iteration codifies Romeo's case study as best practice, the legal fiction becomes policy. If it uses her findings as the empirical basis for a narrower, more conditional version of the designation, the concept survives but loses its present reach.

Either way, the underlying dynamic holds: a policy built for administrative convenience is now setting the terms of legal protection for tens of thousands of people. The evidence from the one case where it has actually been used at scale suggests that the convenience has been doing more work than the safety.

Desk note: The wire framing of "safe third country" tends to treat it as an established legal category with operational refinements. The research above recasts it as a procedural workhorse whose safety function has not been independently verified in the one instance where its use has been large enough to study.