What do children get permanently excluded for?

The recently released school exclusion statistics for 2020/21 saw some significant changes in how the reasons for exclusions were recorded. In order to explain the significance of these changes, and how they are relevant to the debate about exclusions, I will first recap how the data was categorised before those changes.

Since 2015/16, two vague categories – “persistent disruptive behaviour” and “other” – have accounted for the majority of permanent exclusions. (Data from here)

	2015/16	2016/17	2017/18	2018/19	2019/20
Persistent disruptive behaviour (%)	34.6%	35.7%	34.0%	35.2%	34.5%
Other (%)	16.8%	17.6%	18.2%	17.4%	15.7%
Total	51.4%	53.2%	52.2%	52.6%	50.2%

The vagueness of these categories have allowed those campaigning against exclusions to suggest that most exclusions are for relatively trivial matters. “Persistent disruptive behaviour” is often confused with “low level disruption”, although in my experience it more often describes extreme behaviour that is repeated, and the behaviour of those pupils who are completely out of control. Similarly, “other” is assumed to be less serious because it deals with behaviour that is not covered by more serious sounding categories of exclusion. This might have been a reasonable argument if it were not for the fact that bringing offensive weapons into school did not have a category, and might well be included in “other”.

Assaults accounted for around a quarter of permanent exclusions. This was often not given much attention in public debate about exclusions.

	2015/16	2016/17	2017/18	2018/19	2019/20
Physical assault against a pupil (%)	12.3%	13.3%	13.1%	13.3%	12.6%
Physical assault against an adult (%)	10.9%	9.7%	10.7%	10.3%	12.5%
Total	23.2%	22.9%	23.8%	23.7%	25.1%

A further three categories accounted for slightly over a fifth of exclusions. Like assaults, these were not often discussed by those campaigning against exclusions.

	2015/16	2016/17	2017/18	2018/19	2019/20
Drug and alcohol related (%)	7.9%	7.3%	8.1%	8.7%	10.1%
Verbal abuse or threatening behaviour against a pupil (%)	4.7%	4.3%	4.3%	3.8%	3.9%
Verbal abuse or threatening behaviour against an adult (%)	9.0%	8.5%	8.2%	8.2%	7.6%
Total	21.5%	20.1%	20.7%	20.8%	21.6%

Then you have all the smallest categories. Some of these would feature in debates about exclusions; bullying, sexual assault and harassment and racism are big issues. However, opponents of exclusions would often emphasise how few permanent exclusions they account for. Curiously, people who would use these figures to claim we don’t need many permanent exclusions, because sexual assault and harassment, or racial abuse are rare, would immediately pivot to arguing that they are common when discussing those issues in any context other than exclusions.

	2015/16	2016/17	2017/18	2018/19	2019/20
Bullying (%)	0.6%	0.3%	0.4%	0.4%	0.4%
Damage (%)	1.3%	1.2%	1.0%	0.9%	1.3%
Racist abuse (%)	0.2%	0.3%	0.2%	0.2%	0.3%
Sexual misconduct (%)	1.0%	1.3%	1.3%	1.1%	0.8%
Theft (%)	0.6%	0.5%	0.5%	0.5%	0.4%
Total	3.8%	3.7%	3.3%	3.0%	3.0%

For the academic year 2020-2021, changes have been made that address the problem of “persistent disruptive behaviour” and “other” making up most exclusions. Firstly, the category of “other” has been removed. Secondly, a number of new categories have been introduced. Thirdly, it is now possible to give up to three reasons for an exclusion.

The removal of “other” and its replacement with new categories has revealed that “other” does not seem to be minor offences.

	Exclusions	Percentage
Abuse against sexual orientation and gender identity	11	0.28%
Abuse relating to disability	0	0.00%
Inappropriate use of social media or online technology	36	0.92%
Use or threat of use of an offensive weapon or prohibited item	541	13.77%
Wilful and repeated transgression of protective measures in place to protect public health	77	1.96%

I should note that I’ve calculated the number of exclusions in each category as a percentage of the total number of exclusions. The DfE website has not been adequately updated and calculates percentages based on the total number of reasons given for exclusions.

These new categories seem to have been drawn almost entirely from newspaper headlines. The eye opener here is the category “Use or threat of use of an offensive weapon or prohibited item”. When we consider that “other” was the reason for 15.7% of exclusions in 2021, for this new category to include 13.8% of exclusions suggests that the existence of the category “other” has largely been hiding the issue of weapons. I suspect this mainly refers to knives, and I would suggest it is seen as a very pressing reason for exclusion by many schools.

The fact that exclusions can now be given for up to 3 reasons gives us an insight into what was being missed previously. Working out what the new figures reveal is not an easy task. I have done my best, but I am open to any suggestions about how else to approach this. Although there were 3928 exclusions in total for 2020-21, there were 5146 reasons for exclusion given. So there are 31% more reasons than exclusions. This means that we would expect most categories to include proportionately more exclusions than previously. I am going to assume that 31% is the dividing line between those offences that schools were generally willing to give as the main reason for exclusion, and those that were being hidden by the old system of categorising reasons. Generally, I try to avoid using statistics that involve finding a percentage of a percentage, but I can’t think of a better option here. I have calculated each reason as a percentage of the total number of exclusions and the percentage change in the proportion of exclusions in each of the existing categories since 2019-20.

Those which have increased by less than 31% (i.e. the amount by which reasons outnumber exclusions), will presumably be the ones that have not been hidden to a great extent by the old way of giving reasons.

	2019/20	20/21	% Increase
Drug and alcohol related (%)	10.1%	10.4%	2.1%
Persistent disruptive behaviour (%)	34.5%	38.8%	12.6%
Physical assault against an adult (%)	12.5%	14.5%	15.7%
Theft (%)	0.4%	0.5%	28.7%

These are presumably the reasons that tended to be given as the main reason for exclusion, even where a pupil was a repeat offender. I would expect that for the three categories other than persistent disruptive behaviour they were the most serious offence that led to the exclusion. In the case of persistent disruptive behaviour, I suspect that category was probably used to cover repeat offending even where it was more than simply “disruptive”.

The following categories are the ones that were not sufficiently visible when schools could only give one reason.

	2019/20	20/21	% Increase
Verbal abuse or threatening behaviour against a pupil (%)	3.9%	6.7%	75.0%
Physical assault against a pupil (%)	12.6%	22.4%	76.9%
Verbal abuse or threatening behaviour against an adult (%)	7.6%	14.4%	88.1%
Damage (%)	1.3%	2.8%	115.9%
Bullying (%)	0.4%	0.8%	121.7%
Sexual misconduct (%)	0.8%	1.8%	127.8%
Racist abuse (%)	0.3%	1.1%	345.6%

It is noticeable that some of these seem to be quite serious offences and, therefore, it should be a matter of concern that the old system of categorising seemed to understate them.

I had been arguing for a number of years that “other” and “persistent disruptive behaviour” were hiding some serious offences and were not categories for less serious exclusions. I think the evidence here supports that, although I am open to any alternative analyses and explanations of what the figures show. Certainly, anyone who has been arguing that exclusions tended to be for trivial offences now has some explaining to do. In particular, I would draw attention to the figures for assaults. We already knew that around a quarter of exclusions were for assault. Because both types of assault (i.e. against an adult and against a pupil) might be given as reasons for an exclusion, we don’t know how many exclusions in 2020-21 were for assault, but we now know it could be anywhere up to 36.8%.

There’s an even more striking statistic for primary schools. Out of a total of 392 exclusions in primary schools in 2020-21, 203 (51.8%) were for reasons that included assault against an adult. This seems to be quite an important statistic at a time when there have been calls to end primary school exclusions. When you consider that the overwhelming majority of adults working in primary schools are women, the call for greater tolerance for assault against them seems particularly regressive.