Why do some ethnic groups have higher exclusion rates? Part 1

One of the most confused parts of the debate about exclusions from school relates to race and ethnicity. For many journalists, politicians and activists it simply must be the case that ethnic minority pupils, particularly black pupils, are more likely to be excluded. The 2 most recent sets of figures are from the 2019/20 and 2020/21 academic years  They do not show that ethnic minority pupils or black pupils are more likely to be excluded than white British pupils.

	2019/20	2020/21
Ethnicity Minor White British	0.07	0.05
Ethnicity Major Black Total	0.07	0.05
Ethnicity Minority ethnic pupil	0.05	0.04

“Minor” here means a category that cannot be subdivided further; “major” is one that can be subdivided. The rates here are the number of exclusions for the pupils in a category, divided by the number of pupils in the category, written as a percentage, so 0.05 means 1 in 2000 pupils were excluded during the year. The DfE rounds exclusion rates to two decimal places. If we calculated further decimal places, the exclusion rate for black pupils would actually be lower than for white British pupils. However, in both years there were over 100 excluded pupils who were categorised as “ethnicity unclassified” which gives us a potentially large margin of error for these statistics, so comparisons based on only those more precise figures probably can’t be relied upon.

Why are so many people convinced that exclusion rates for ethnic minority pupils, and black pupils in particular, are higher? One reason is that this was the case historically. As recently as the 00s, ethnic minority pupils were slightly likelier to be excluded than white British pupils and black pupils were around twice as likely to be excluded. It is only recent years that ethnic minority pupils have become much less likely to be excluded than white British pupils, and it is only in the last two years that black pupils were no more likely to be excluded than white British pupils.

(Data from here. Despite what I said above about the margin for error, I have calculated the figures as precisely as possible just to make the lines clearer on the graph.)

I have in the past suspected that another reason for the widespread belief that black pupils are more likely to be excluded is because so much coverage of exclusions is driven by London based media and politicians. Racial disparities exist in London that don’t exist elsewhere. Inner London has a really low rate of permanent exclusions for white British pupils. In 2019-2020 London was the only region of England where the ethnic minority exclusion rate was higher than the white British exclusion rate. And in that year, while there were other regions where the exclusion rate for black pupils were a little higher than for white British pupils, Inner London was the only region where it is a lot higher.

(Data from here).

In 2020-21, however, this did change slightly, with the disparity between black and white British pupils narrowing dramatically in Inner London and widening in the South West.

However, Inner London remains the only region where ethnic minority pupils are more likely to be excluded than white British pupils.

For whatever reason, there are countless commentators who simply cannot admit what recent figures show about racial disparities. I am continually seeing false claims about this topic. Often this is justified by using out of date data. Sometime people use white pupils (including ethnic minority pupils) as the baseline for comparison instead of white British pupils, which creates a small disparity in 2019/20. Sometimes people just lie or repeat something they’ve heard without checking if it’s true.

However, the single most common way to suggest that exclusions discriminate is to subdivide into much smaller categories than ethnic minority pupils or black pupils, then pretend that what is true for some of these smaller subgroups is true for much larger groups. Exclusion data is collected for 18 separate “minor” groups (i.e. groups that cannot be further subdivided):

• Ethnicity Minor Any Other Ethnic Group
• Ethnicity Minor Any other Asian background
• Ethnicity Minor Any other Mixed background
• Ethnicity Minor Any other black background
• Ethnicity Minor Any other white background
• Ethnicity Minor Bangladeshi
• Ethnicity Minor Black African
• Ethnicity Minor Black Caribbean
• Ethnicity Minor Chinese
• Ethnicity Minor Gypsy Roma
• Ethnicity Minor Indian
• Ethnicity Minor Irish
• Ethnicity Minor Pakistani
• Ethnicity Minor Traveller of Irish heritage
• Ethnicity Minor White British
• Ethnicity Minor White and Asian
• Ethnicity Minor White and Black African
• Ethnicity Minor White and Black Caribbean

Not surprisingly, these categories do not all have the exact same exclusion rates. There are even a few persistent inequalities. We can divide the categories into 3 separate groups.

Those that consistently have a lower exclusion rate than white British pupils. These have all had an exclusion rate that (to two decimal places) was lower than white British pupils in every year of the last 5 years.

• Ethnicity Minor Any Other Ethnic Group
• Ethnicity Minor Any other Asian background
• Ethnicity Minor Any other white background
• Ethnicity Minor Bangladeshi
• Ethnicity Minor Black African
• Ethnicity Minor Chinese
• Ethnicity Minor Indian
• Ethnicity Minor Pakistani
• Ethnicity Minor White and Asian

Those that consistently have a higher exclusion rate than white British pupils. These have all had an exclusion rate that was higher than white British pupils in every year of the last 5 years.

• Ethnicity Minor Any other black background
• Ethnicity Minor Black Caribbean
• Ethnicity Minor Gypsy Roma
• Ethnicity Minor Traveller of Irish heritage
• Ethnicity Minor White and Black Caribbean

Those where the exclusion rate is not consistently above or below the rate for white British.

• Ethnicity Minor Any other Mixed background
• Ethnicity Minor Irish
• Ethnicity Minor White and Black African

A list of exclusion rates for each ethnic group for any of the last five years will therefore, put white British pupils somewhere between the 6th and the 9th place for exclusions out of 18 ethnic groups. This means that, at a glance, it white British pupils are around halfway down the list and, therefore, it can be claimed that up to half of ethnic minority groups are disproportionately excluded. However, this can be misleading because the 18 ethnic groups are very different in size. When you take account of how many pupils there are in each of these groups, the ethnic groups that have a consistently higher exclusion rate than white British pupils account for just 11% of England’s ethnic minority pupils.

Consistently lower than white British pupils	2 176 731
Consistently higher than white British pupils	307 959
Not consistently above or below white British pupils	293 451

Much of the commentary we encounter consistently cherrypicks the ethnic groups in the orange slice and implies that they represent the whole pie chart.

It is unlikely that the groups with a higher exclusion rate would be so small just by chance. Ethnic minority pupils have been subdivided so that there is a moderate negative correlation (r=-0.44) between an ethnic group’s exclusion rate and its size, i.e. the smaller an ethnic group is, the more likely it is to have a high exclusion rate.

The categories for ethnic groups seem to be based on those used in the census. From what I’ve read the development of categories over time involves quite a lot of consultation and many considerations and I wouldn’t claim to understand the details. However, it would appear that there is a greater willingness to split up disadvantaged groups (like Travellers) into smaller groups, while keeping other, less disadvantaged groups (like the category made up of half a million white ethnic minority pupils) undivided. And while this may make sense for monitoring disadvantage, it means that lists based on this division give undue prominence to some small, but disadvantaged, groups. And that has enabled statistics describing just 11% of ethnic minority pupils to completely dominate the debate about exclusions and ethnicity. So, my first answer to the question of why some ethnic groups have higher exclusion rates is that it’s partly because of how the data has been sliced up, and that’s just a feature of how we analyse exclusions, not a feature of the exclusions themselves.

Of course, even when we recognise that the 11% of ethnic minority pupils are unrepresentative of ethnic minority pupils in general, that doesn’t mean they shouldn’t be an important consideration. But it does mean that in finding out why they have higher exclusion rates, we may need to work out what makes them different from the vast majority of ethnic minority pupils. This is a potentially controversial question that I hope to return to.

Advertisement