Mail's reporting of crime statistics questioned

In January, the Mail published this article on the August riots:

The UK Statistics Authority released a Monitoring Brief (pdf) on 20 March which looked into the Mail's article and concluded:

the reporting of this information by the Daily Mail is likely to have left its readers with the impression that far fewer crimes were recorded as a result of the disorder in August than was actually the case.

The Mail's article - by Jack Doyle and Graham Smith - began:

The riots that left whole neighbourhoods up and down the country in a state of ruin last August were the worst civil disturbances for a generation. But reading crime figures released yesterday, it is almost as if the five days of widespread looting and violence never took place.

But the UK Statistics Authority said:

The Daily Mail article quoted the correct number of specific offences of disorder recorded by the police, but did not give the numbers of the other offences that it used to illustrate the disorder in each area. These included serious violent offences (such as murder), criminal damage (e.g. to buildings, cars and arson offences), and acquisitive crimes (such as burglary, robbery, vehicle and other theft).

And:

The Daily Mail went on to correctly note, but question, the way in which the Home Office Counting Rules govern the recording of riot, public order offences and violent disorder by police forces. In some cases the Mail’s reporting is likely to have left readers with the impression that far fewer crimes were recorded as a result of the events in August than was actually the case. For example, the article says that in Croydon the Metropolitan Police only recorded 7 disorder offences, while in fact a total of 430 offences were recorded. The total number of offences that were recorded by the ten police forces that experienced more extensive disorder (recording 20 or more disorder-related offences) was 5,112. This total includes the 141 specific offences of disorder, as well as more than 2,500 acquisitive crimes, 1,800 offences of criminal damage and 360 violent offences.

(Hat-tip to journalism.co.uk)

Fun with numbers

One article from the Mail and one from the Sun today which both show a less than comprehensive grasp of statistics. But it's not solely down to stupidity - the headlines they have used fit their agenda, even if the numbers don't.

So from the Mail:

'Winning banned in two-thirds of schools'. And the article does reveal a new poll that:

surveyed almost 300 primary and secondary schools and found that 69 per cent reward all participants in sports days.

It doesn't explain what these rewards are, but that doesn't necessarily mean 'winning' is 'banned'.

Indeed, it then goes on to totally contradict the headline:

Nine per cent of all schools refuse to single out any winners at all.

Ah. So it's only 9% then. Not two-thirds.

Over in the Sun, a report on a poll about a referendum on changing the voting system. YouGov asked:

In view of spending cuts, is it appropriate to spend £80million on an AV referendum now?

It's not clear where the £80million figure comes from - the article states it could cost 'up to' that amount. And while 46% said inappropriate, 35% said appropriate and 19% said don't know.

But they also asked:

Regardless of how you would vote, do you support the principle of holding a referendum?

69% said yes and only 12% were opposed.

The Sun's headline for this support for a referendum?


The Sun oppose the referendum, as the Mail wants everyone to think political correctness has gone mad at school sports days, hence the headlines. Presumably they just hope people won't notice what's actually been said.

(Hat-tip Adam Bienkov)