Are exclusion rates for black Caribbean students up to six times higher?

The Guardian recently chose to print an utterly uncritical piece on No More Exclusions, the campaigning group that wants even the most dangerous young criminals to stay in school with their victims. More accurate information about that group and their appalling views can be found here (from me) or here (from the TES). While I’m very disappointed at the Guardian for this, in this post I wish only to address one point made in that article:

The racial disparities are stark, with exclusion rates for black Caribbean students up to six times higher than those of their white peers in some local authorities, according to a recent Guardian analysis.

I wrote about that “analysis” when it was published, in this post: How misleading was today’s Guardian article on exclusions?  

The entire article was cherrypicked data, almost certainly intended to mislead. In particular:

• Ethnic minority groups were broken down into the smallest possible categories and those ethnic groups that had lower exclusion rates than white British pupils were ignored (despite being the overwhelming majority of ethnic minority pupils);
• Fixed term exclusion rates were used instead of permanent exclusion rates;
• England was broken down into 150 LAs, and figures were given for the LAs with the most extreme discrepancies.

The article then repeated claims about how this was an incredible injustice for pupils from an ethnic minority background, and calls for changes in national policy, as if the statistics applied to all ethnic minorities in the whole of England. Later Guardian stories (like this one and, before it was corrected two weeks later, this one) referred back to the article as if it was about black pupils in general rather than black Caribbean pupils in a handful of LAs.

However, even if the article had been honest in the first place, there are several reasons why the specific statistic about black Caribbean pupils being up to six times as likely to be excluded should not still be being repeated.

Firstly, it is not, as described, “a recent Guardian analysis”. It was published over a year ago. When I wrote about that in my previous blogpost, I assumed it was based on data that had been released in February 2021 and it had taken the Guardian a month to subdivide the data in ways that supported a narrative of discrimination. Looking back now, I can’t actually see any reason to think it wasn’t based on data that was first published in July 2020. Regardless, it was data about the 2018-19 school year, and data on the 2019-20 school year was published in July 2021 and so, far from being “recent”, that analysis has been out of date for 10 months.

Secondly, in the more recent statistics, the term “suspensions” has replaced “Fixed Term Exclusions”. Suspensions are when a student is temporarily removed from school, usually for a day or two and rarely for more than a week. When a child is expelled it is called a “permanent exclusion”. Referring to suspensions as exclusions has allowed anti-exclusion campaigners to misrepresent the data for many years by confusing expulsion and suspension. As mentioned earlier, the original Guardian “analysis” used the figures for Fixed Term Exclusions. While it is in a paragraph that mentions suspensions, to continue to quote it in an article about exclusions, where many of the claims are clearly about permanent exclusions, is misleading.

Thirdly, although the “up to six times higher” statistic was apparently the headline when that article was the cover story of the print edition of the Guardian, the online version had a different headline. In the text it was admitted that the six times figure came from Cambridgeshire, and that  “Cambridgeshire has a relatively small number of Caribbean students, which partially explains the disparity”. The new headline was “Exclusion rates five times higher for black Caribbean pupils in parts of England”. So in it’s article about No More Exclusions, The Guardian is quoting something that is not just out of date, but that they have previously corrected.

That last point might seem relatively minor, after all, does the difference between six times and five times matter? In fact, the Guardian reported that Brent was excluding black Caribbean pupils at 5.9 times the exclusion rate for white pupils. So what difference does it make that the six times figure is marginally higher? I think what’s most misleading about including it is that in mentioning Cambridgeshire, it suggests the disparities being described are a national issue. Without the Cambridgeshire figure, every single LA featured in the Guardian’s graphic (below) is in London.

London does contain the majority of England’s black Caribbean pupils, so it is perhaps unsurprising that many would be from London. However, if you look at the racial disparities in exclusions between black Caribbean and white British pupils, while racial disparities exist outside of London, they seem to be less common and on a smaller scale. I have looked at LAs with more that 750 black Caribbean pupils, a somewhat higher threshold than the Guardian’s. This is what the disparities look like (London LAs are in blue, non-London LAs are in orange).

 

I don’t mean to suggest that disproportionate suspension of black Caribbean pupils is not an issue. or even that it is never an issue outside of London. But, it is a far more pressing issue in London, where most of England’s black Caribbean pupils go to school and where many LAs have very high exclusion rates for black Caribbean pupils. At some point I hope to blog in detail about how London’s exclusion data is very different to the rest of England’s, but here I hope it is obvious how the Guardian’s methodology distorts the exclusions debate to make one of London’s issue seem like a more general issue for England.

If we accept that it is mainly London that has a problem with disproportionate suspension of black Caribbean pupils, then we might end up looking at what else is exceptional about suspensions in London. We would then, perhaps be interested in whether suspension rates varied between regions.

From the above graph, the most noticeable thing about suspensions in London is how rarely they are used. Correlation is not causation, so I’m not going to claim that having a low use of suspensions widens racial disparities. But this should be reason to stop assuming that racial disparities are a reason to use suspensions less. The lesson from London could well be that where people let kids off of suspensions that are deserved, it is white British kids who are most likely to be let off. At the very least, we should consider the possibility that it is not just sanctions that might have a disproportionate impact on some ethnic groups. It might be that reluctance to use a sanction, or political pressure not to use it, has a disproportionate impact on different ethnic groups. The Guardian was so busy looking for reasons to argue against exclusions, that it missed the real story of the suspension data. The story that London has a unique problem with racial disparities in suspensions, and that this has happened in a city that stands out for having kept suspensions low.