January 20, 2025 - 1:00pm

In an article published last week, Conservative MP Neil O’Brien presented data for England and Wales showing that the proportion of convictions for which the ethnicity of the defendant was not recorded has been rising over time — from around 10% in 2011 to roughly 34% in 2024. He also noted that rises in the proportion of convictions where the age or sex of the defendant was not recorded have been much smaller. The rise in missing ethnicity “trashes the statistics as a whole”, according to O’Brien. But what explains this trend?

One hypothesis is that certain police forces and courts have become less likely to record all defendants’ ethnicities, for whatever reason. An alternative hypothesis is that they have become less likely to record the ethnicities of defendants from minority backgrounds — perhaps in order to conceal the scale of crime committed by these groups. The possibility that they have become less likely to record the ethnicities of white defendants seems somewhat remote.

If the alternative hypothesis is true, ethnic minorities should be overrepresented among defendants with missing ethnicity. However, two pieces of evidence suggest that ethnic distribution of defendants with missing ethnicity is similar to the ethnic distribution of those with known ethnicity.

The first is one that I have referenced previously. The Government publishes an ethnic breakdown of the prison population, and here the number with missing ethnicity is only 1%. If defendants from minority groups are less likely to have their ethnicity recorded, they should be comparatively overrepresented in the prison population. However, there is no evidence of this. The ethnic distribution in prison closely matches the ethnic distribution of prosecutions for indictable-only offences. As O’Brien noted in his article, the rise in missing ethnicity is present for prosecutions as well as convictions.

As for the second piece of evidence, the year with the lowest percentage of defendants with missing ethnicity in O’Brien’s data is 2011. If defendants from ethnic minorities are less likely to have their ethnicity recorded, they should be comparatively overrepresented in the 2011 figures, compared to more recent figures from 2022. Once again, however, there is no evidence of this.

Prosecutions data by ethnicity shows no significant change over time
Defendants by ethnic background, 2011 vs. 2022

To properly compare the two sets of figures, we need to adjust for the fact that the white population has become smaller while the non-white population has grown in the same period. Once we do so, there is a close match between the two. In the adjusted 2011 figures 23.2% of defendants were from minority backgrounds, compared to 23% in the more recent figures.

For the sake of simplicity, I adjusted for demographic change using figures for the overall population. Another caveat is that Chinese Britons were classified as “Other” in 2011 but are now classified as “Asian”. (They commit very little crime so this should barely affect the numbers for the two relevant categories.)

The rise in the proportion of defendants with missing ethnicity is certainly concerning, and O’Brien is right to draw attention to it. However, evidence suggests the ethnic distribution of those with missing ethnicity is similar to that of those with known ethnicity, in which case published figures on crime and ethnicity are largely unaffected.


Noah Carl is an independent researcher and writer.

NoahCarl90