Governments should establish individual safety before deporting migrants to third countries

What's this about?

People disagree about whether governments should check each migrant’s safety before sending them to another country.

The key question is whether that person may face serious harm there.

What supporters say

• A no-return rule says governments must not send people toward harm, such as torture or unfair abuse.
• Sending someone to a third country may bring new dangers, such as jail, poor care, or removal to danger.
• A country may seem safe overall but still be unsafe for children, sick people, or LGBTQ people.

What critics say

• Legal transfers can let more than one country share the duty to protect people.
• Governments may use transfers to control unsafe or unplanned border crossings.
• Checking every person’s case can take more time, staff, and money.

How to read this

The number of points on each side does not show who is right; stronger proof matters more than a longer list.

The bottom line

The evidence strongly supports checking each person’s safety before deportation.

Governments may still transfer people when real, reliable protection works in practice.

The claim is that governments should determine whether a third country is safe for a particular migrant before carrying out a deportation. The evidence strongly supports that approach, while still allowing transfers when reliable protections are demonstrated in practice.

The case for

The strongest argument is legal. The principle of non-refoulement bars governments from sending people to places where they face persecution, torture or other serious harm. That requires a forward-looking assessment of the person’s actual risk, not simply reliance on a general “safe country” label. Article 33 of the Refugee Convention and Article 3 of the Convention against Torture both focus on the foreseeable danger to the individual. 1

European human-rights law has similarly stressed the need to examine reception conditions, access to asylum procedures and the risk that a receiving country could send someone onward to danger. A diplomatic promise or country classification cannot automatically replace effective scrutiny.

A transfer can also create risks that did not exist, or were not assessed, in the original destination. Migrants may face detention, poor reception conditions, lack of legal status, an inability to seek protection or onward removal. Research on Australia’s offshore processing system has documented serious psychological and physical harm linked to prolonged stays offshore. Distance and formal admission by another government do not, by themselves, prove safety (see Figure 1). 2

General country assessments can also miss risks tied to a person’s circumstances. UNHCR guidance points to admission, protection from refoulement, a fair asylum process, adequate reception, legal status and basic rights as relevant conditions. Individual review is especially important for children, trafficking survivors, LGBTQ people, people with disabilities and those with medical or family needs. 3

Prior review also gives migrants a meaningful chance to challenge both the transfer and the treatment they may receive after arrival. Rapid or automatic removals can make mistakes difficult to correct once a person has left the country.

The case against

Governments have legitimate reasons to use carefully designed third-country transfers. A transfer can lawfully share responsibility for protection if the receiving country admits the person, provides a genuine and fair asylum process, respects non-refoulement and offers basic rights in practice. Safety does not mean proving that a country is risk-free; it means showing that protection is real and sufficiently reliable for the proposed transfer. 4

Governments also argue that transfers can help manage irregular migration. But the evidence for a broad deterrent effect is much weaker than the legal evidence supporting protection safeguards. Studies of offshore processing have found it difficult to separate deterrence from other policy changes, while also documenting substantial humanitarian costs. They do not show that safety checks can be dropped in pursuit of deterrence. 5

Individual assessments can require legal assistance, country expertise, appeals and ongoing monitoring. That may slow removals and increase costs. This is a practical concern, but it shows the burden of implementing safeguards—not that transfers are safe without them. 6

The bottom line

The evidence strongly favours a qualified requirement for prior, individualized safety assessment. Governments should consider current country conditions and the migrant’s personal circumstances before deportation, and should require genuine admission, access to an effective asylum process and appeal, protection against onward removal, and monitoring after transfer.

This is not an absolute ban on third-country deportations. Lawful transfers remain possible when those protections are demonstrated and remain enforceable. Diplomatic assurances can help, but they are not conclusive: the Rwanda judgment showed why promises must be tested against institutional capacity, real procedures and the risk of refoulement.

The main uncertainty is how officials can reliably assess changing conditions and predict whether safeguards will work for different people. Still, the protection-based case is backed by stronger legal and institutional evidence than the arguments for avoiding individualized review. Administrative and responsibility-sharing interests matter, but they cannot substitute for showing that the individual will be admitted, protected and able to obtain effective protection in practice.