Are assaults on police officers rising?

“Shocking rise in police assaults: 28 attacks a day on officers in crime epidemic,”

Daily Express front page, 13 August 2019

On Tuesday statistics suggesting a rise in assaults on police officers made the front page of the Daily Express, and were also reported elsewhere.

It’s true that all recorded assaults on officers rose by 18% in the 12 months to March 2019 compared to the preceding year.

But due to a major change in the way the data is collected, it’s not possible to compare the complete statistics for this year with those of any year before 2017—so it’s not possible to say with any certainty what’s happened over the past four years, as the Express did.

There are some additional issues with the reliability of the data: the Home Office has said the numbers “are not thought to provide a complete picture of assaults against police officers”.

The claim that officers are being injured at a rate of 28 attacks per day is an accurate extrapolation from the available data, but given this data’s limitations it is likely to underestimate the number of these assaults.

It’s not possible to reliably compare 2018/19 with four years ago

The Office for National Statistics (ONS) collects two types of crime data on police officer assaults for England & Wales (figures for Scotland and Northern Ireland are separate). These types are “assault without injury on a constable” and “assault with injury on a constable”.

In 2018/19, there were 20,578 recorded incidents of the first type, a rise of 13% compared to the previous year, and 10,399 of the second type, up 27% on the year before.

Together, this meant there were 30,977 recorded crimes involving assaults on a police constable, an 18% increase on the 26,295 that were recorded in 2017/18.

However, the Express story based its headline claim on a comparison of this year’s data with that of 2015/16, claiming assaults had “risen by a third in just four years”.

Specifically, the newspaper claimed that: “Bobbies were victims of 10,399 assaults that caused injuries last year – up 32 per cent from 7,903 recorded in 2015/16.”

But you can’t compare those two numbers.

Before April 2017, “assault with injury on a constable” was not a crime category, and data on these incidents was collected via officers self-reporting their injuries to their forces’ HR or health and safety systems.

Since April 2017, a separate crime category for these incidents was set up, and so the last two years have seen the assaults recorded as specific crimes.

There are differences between what was reported in the health and safety systems and what is now recorded as a crime—particularly because it was usually not compulsory for officers to self-report cases. That means it was likely to be under-counting the number of assaults, which would make the rise between then and now look bigger.

The same problem applies to comparing all recorded assaults, including those that did and didn’t result in injuries.

There’s actually another way of arriving at the Express’s figure, which is to track the change in the number of assaults without injury on a constable. This can be reliably compared across the last four years, because there have been no changes to the way incidents are recorded.

The number of these assaults has risen by 33% since 2015/16. The problem, though, is that this comparison does not include all recorded assaults, and excludes the most serious ones. 

The only meaningful comparison we can make for measuring recorded assaults on police officers is between 2017/18 and 2018/19.

The ‘28 injuries per day’ figure

Another claim made by the Express in its coverage was to estimate an average of 28 attacks per day resulting in injuries to police officers. The Daily Mail repeated this claim.

As there were 10,399 recorded assaults with injury to a constable in the year to March 2019, and that equates to 28 per day, this is a reasonable claim. But it doesn’t give a complete picture.

The most serious assaults can be categorised under such crime categories as “attempted murder”, which means the victim’s status as a police officer is not recorded for these cases and so they’re not included in the police officer assault statistics.

This means the ‘28 injuries per day’ figure is likely to be an underestimate.

It’s also worth noting the Express’ claim that officers are being injured at a rate of 28 assaults a day while “on duty” isn’t quite right, because the available data does not distinguish whether or not an officer was on duty when the assault occurred.