A quick one from me on the stats (specifically on Home Office decision making quality in asylum claims)

I decided this was too long to include in tomorrow’s post, so have done it as a short standalone here. Today, the Home Office has started publishing statistics on asylum decision quality again.

In previous statistics releases, the Home Office provided just a single “Decision quality” figure. So for 2023/24 all we had was “52%”. In October 2022 the Home Office introduced a new method of measuring quality and this stats release is the first time that they have reported on this new method.

I don’t love it, as I think it raises quite a lot of questions, but at least we actually have some data again, and it is useful to have a better idea of what the Home Office considers to be a good enough decision to go out.

Now we have a breakdown of five separate categories – “Quality scores” 1 through 5. QS 1 and 2 relate to minor errors (the latter with 20% or more of these), defined as “An error which does not detract from the consideration and would not affect the outcome of the decision, and should be quickly rectified”. QS3 means that a decision has a “significant error” defined as “An error which detracts from the quality of the consideration of the decision and requires attention to address serious weaknesses or omissions.”

QS4 and QS5 contain “fail errors”, but decisions categorised as QS4 are still deemed to be “probably correct”. Decisions scored QS1 through to QS4 are considered “correct” (although those scored QS3 and QS4 will be returned to the decision maker to correct). Decisions with a score of QS5 are considered “incorrect”. The results are below:

So of the above, in each of the past two years the Home Office has only deemed 6% of the decisions that they have looked at as “probably wrong or involving special policies”. This seems unrealistically generous given how many of the Home Office’s decisions are sustained following appeal.

The tribunal statistics for July to September 2025 showed that of the asylum appeals concluded during this period, 49% had a decision made by the tribunal and of those, 39% of asylum appeals succeeded.

A huge 37% of asylum appeals were concluded by being withdrawn, either by the appellant or by the respondent. There is little incentive for appellants to withdraw their appeals and it seems reasonable to conclude that a significant proportion of those were decisions taken by the Home Office that the refusal was unsustainable.

The reasons for those withdrawals is really important, not least when trying to assess the full extent to which the Home Office is making unlawful decisions, but it is not currently published. Even without this data, that 6% internal fail rate of decisions indicates that the Home Office has a significant problem with assessing the quality of its own decisions.

Tribunals data for all of 2025 will be published on 12 March 2026.