One of the changes wrought by the Nationality and Borders Act 2022 (remember that?) is an apparent increase in the standard of proof in refugee status claims. This change applies to everyone who applied for asylum on or after 28 June 2022. There has been a huge waiting time for initial decisions and there’s also a growing waiting time for appeals. But some of these asylum claims are finally starting to filter through to judges.

So, what might be the effect of the change, in theory and in practice?

What was the refugee status standard of proof before?

Previously there was a relatively simple and straightforward standard of proof: is there a reasonable degree of likelihood that the person in question will be persecuted on their return to their country of origin? This could be expressed with different formulae of words. Is there a real risk of serious harm? Are there substantial grounds for believing? Is the fear well founded? All of these were considered to mean the same thing.

Exactly what that actually means in real life is, frankly, not so simple. It cannot be expressed as a percentage possibility. It’s perhaps something akin to “non-negligible” or “beyond fanciful”. The consequences of a wrong decision are very, very bad, after all. The person concerned might be tortured or killed. When it comes down to it, the assessment is probably best understood as one that is socialised into lawyers, decision-makers and judges through training and experience. I’ll come back to that point at the end.

In any event, what we do know for certain is that the ‘reasonable degree of likelihood’ test is supposed to be less than the normal civil standard of the balance of probabilities. The balance of probabilities is easily understood: is something more likely than not? Reasonable degree of likelihood is less probable than that.

We also know — or knew — that the standard of proof was applied to everything about an asylum claim. Was it reasonably likely the person was telling the truth, that the events happened as claimed, that there is a risk of sufficiently bad things happening in future? In the early days of refugee law, some had argued that the civil standard of the balance of probabilities should be applied to the assessment of past facts and the reasonable degree of likelihood just to the assessment of future risk. Judges eventually held in a case called Karanakaran that the correct approach was to apply the same standard to everything.

The law since then had been settled, static and everyone knew what they were doing.

What is the refugee status standard of proof now?

The Nationality and Borders Act 2022 requires decision-makers and judges, when assessing a refugee status claim, to adopt the split approach that some preferred back in the 1990s.

First of all, section 32(1) says:

In deciding for the purposes of Article 1(A)(2) of the Refugee Convention whether an asylum seeker’s fear of persecution is well-founded, the following approach is to be taken.

So, what follows applies to refugee status claims. It does not apply to human rights claims based on Article 3 of the European Convention on Human Rights. Nor is there any equivalent applied to human rights claims anywhere else in the legislation; to impose a different standard of proof in human rights claims would be incompatible with the very well established jurisprudence of the European Court of Human Rights. Not only that but it would also be incompatible with all of the other human rights conventions, like the International Covenant on Civil and Political Rights, the Convention Against Torture and so on, which follow the same approach.

Section 32(2) then says this:

The decision-maker must first determine, on the balance of probabilities—

(a) whether the asylum seeker has a characteristic which could cause them to fear persecution for reasons of race, religion, nationality, membership of a particular social group or political opinion (or has such a characteristic attributed to them by an actor of persecution), and

(b) whether the asylum seeker does in fact fear such persecution in their country of nationality (or in a case where they do not have a nationality, the country of their former habitual residence) as a result of that characteristic…

If the asylum seeker does have a relevant characteristic and fears persecution, then section 32(4) then says this:

The decision-maker must determine whether there is a reasonable likelihood that, if the asylum seeker were returned to their country of nationality (or in a case where they do not have a nationality, the country of their former habitual residence)—

(a) they would be persecuted as a result of the characteristic mentioned in subsection (2)(a), and

(b) they would not be protected as mentioned in section 34.

The point about this approach only applying in refugee status claims is reinforced by section 30, which states to whom these parts of the Act apply and for what purpose: an any person, court or tribunal determining whether a person is a refugee within the meaning of Article 1(A)(2) of the Refugee Convention.

Do refugees now need to demonstrate a real subjective fear?

Another issue arises. Does a person have to experience “fear” in the sense of an emotion akin to dread or terror in order to be a refugee? Or does “fear” in this context mean apprehension or expectation that something may occur? Or is the real question just one of objective risk: what will really happen to the person in the future if they are sent into that situation?

This will not matter in most cases. But it might matter for children, those who lack mental capacity or, conceivably, those who are particularly courageous. What happens where a child does not understand the danger they are in, for example? They have no subjective “fear” as such but there might nonetheless be a real risk of being persecuted.

There are authorities both ways on this question but in practice decision-makers and judges have not interrogated the emotional state of the applicant but rather the degree of objective risk. Might the Nationality and Borders Act 2022 force an emotional inquisition?

It might. Section 32(2)(b) requires the decision-maker to assess “whether the asylum seeker does in fact fear such persecution in their country of nationality”. Separately, section 32(4) requires the decision-maker then to assess whether “they would be persecuted as a result”. Separating the assessments might be taken to require separate assessments. If so, that’s a retrograde step that, in theory, can only harm vulnerable would-be refugees like children.

It barely matters in the end. That split approach would not apply in a parallel human rights claim anyway, which requires an objective assessment based on risk not subjective emotional state.

Why have a split standard of proof which only applies in refugee status claims?

I don’t know. I can’t think of any reason why this is useful to anyone. It’s a total waste of everyone’s time. It complicates and slows down decision-making and in legal terms achieves no ultimate outcome whatsoever. Anyone denied refugee status on this approach should succeed on human rights grounds anyway.

On the face of it, what we should expect is that decision makers now have to separately assess refugee status claims and human rights claims according to different approaches to the standard of proof. That would increase the complexity of the process and the length of written decisions. The effect, in theory, would be that some people who would otherwise have been recognised as refugees will now not be so recognised. But the same people would still succeed on factually identical human rights claims in which the old approach continues to apply.

There is a line of thinking that refugee law has been rendered essentially redundant by human rights law, which applies to all the same situations as refugee law and some extra situations too. In a Venn diagram of protection claims, the circle of refugee law claims has arguably already been entirely gobbled by the human rights one. This development is grist to that mill. I’m not going to get into that debate here other than to say that international refugee law does include concrete rights for refugees in a way that international human rights law does not.

But in the UK, right now, there’s essentially no substantive difference between the status or rights awarded to a person recognised as a refugee and a person who succeeds on human rights grounds. So the change to the standard of proof in refugee status claims makes no real difference if the same person goes on to succeed on human rights grounds, as they should. There’s been so much nonsense in the last few years I’m not sure I’m remembering this correctly, but at one point I think the Home Office might have decided to make human rights protection less generous than refugee protection by requiring a longer period of residence until settlement was achieved. But I think that was dropped with the abandonment of the ‘differential treatment’ policy. That policy was dropped because it simply created more work for Home Office officials, had no deterrent effect and there was the matter of the massive asylum backlog to focus on.

I think I’m right in saying that the period of stay granted is the same, so is the residence period before qualifying for settlement and so are family reunion rights. It is harder for those with human rights protection to get a travel document for travel abroad.

What will really happen in practice?

We’ve already seen one tribunal case on the new approach in refugee cases: JCK v Secretary of State for the Home Department (Botswana)[2024] UKUT 100 (IAC). We did a short write-up of the case here: Upper Tribunal gives guidance on assessing well-founded fear under the Nationality and Borders Act. Weirdly, this point about the new standard of proof only applying in refugee claims is not made anywhere in the determination. It did not need to be as was not directly relevant to the decision in question but I ask, rhetorically, when did that ever stop an Upper Tribunal Immigration and Asylum Chamber judge from opining on something?

The first possibility is the one I’ve outlined above. Decision-making takes longer but there’s no real change in outcomes other than a very small fall in grants of refugee status matched by an exactly equivalent small rise in grants of human rights protection. We probably would not be able to see the change in the statistics because (a) it would probably be small and (b) it would be impossible to separate from other factors causing a changing grant rate.

The second possibility is that everyone basically ignores the substance of the change while having also to insert some extra paragraphs and reasoning into decisions. I cannot imagine experienced decision makers saying to themselves “I’d have granted refugee status on this one previously but now I’m going to deny it even though there’s a real risk what they say happened to them is true”. Not that this would cause the person to be sent back because they’d stay on human rights grounds anyway.

The final and third possibility is that the change, while legally and theoretically nonsensical, nevertheless has an impact on decision-making culture. The change might well be a hook from which to hang new training and supervision of officials at the Home Office. It might lead to an injection of additional scepticism. That might lead to more refusals of refugee claims AND human rights claims too. The turnover and also expansion of staff at the Home Office is such that the number of experienced decision-makers must be tiny, so the inculcation of a new decision-making culture certainly seems plausible.

We are seeing the grant rate drop at the moment. It was at a high of 75% but was down to 66% last time I looked. Which is still high, of course. I’ve been rationalising this as an outcome from the bringing-forward of high grant-rate nationalities to the front of the long queue. Once all of those are granted, a far higher proportion of lower grant-rate nationalities remain still to be decided and we would therefore expect the grant rate to fall.

But maybe, with the rapid expansion of decision-makers from 600 to 2,400 and the new training the new officials have received, there is more going on than that.

As an aside, rejection of asylum claims creates massive additional work for the state. Appeals need to be processed, which involves judges, hearing rooms, Home Office lawyers and claimant lawyers. Asylum seekers need to be supported through the appeal process. In theory, those whose applications and appeals fail are supposed to be removed. Granting status avoids all of that extra work and expense and would allow the Home Office to focus on other stuff.

That is just the Home Office, though. It is far harder to displace judicial decision-making culture, which is likely to be far more durable. If so, and Home Office refusals rise, we would expect the appeal grant rate to rise. It’s worth keeping an eye on.

Interested in refugee law? You might like Colin's book, imaginatively called "Refugee Law" and published by Bristol University Press.

Communicating important legal concepts in an approachable way, this is an essential guide for students, lawyers and non-specialists alike.

Buy now