Issues & Ideas | 23rd April 2015

Police crime figures haven’t fallen – but that’s a good thing

Is crime down or up? The figures that matter shows it’s down, and the police may just be getting better at reporting real stats.

Photo: Getty


“I mean the election’s fucking over, right? Who are we doing this for?”

One of the recurring themes of The Wire, the landmark 2000s-era television series on Baltimore’s drug-addled plight, is how the police fix crime statistics. “Making robberies into larcenies, making rapes disappear… you juke the stats and majors become colonels.”

Do our police fix the stats? Last summer, when this site was preparing to launch, the ONS released its latest crime figures for the UK. Those figures showed that their measure of crime – a survey-based measure that they began in the early 1980s – was reporting a significant fall in the crime rate,  while the police’s official figures were showing a negligible, 0.1 per cent fall.

“You juke the stats and majors become colonels.”

Our view was that the police had potentially simply started to clean up their stats, after years of ‘juking the numbers’ in order to report falls in crime. In January last year the government’s top statistician downgraded the police’s self-reported numbers – they are no longer counted as a ‘national statistic’ – after whistleblowers told a select committee of MPs two months earlier that police often fiddled with the numbers.

Today, the latest numbers have been released, and tell a very similar story to the one they told last summer. The ONS’ survey measure of crime has reported a 7 per cent fall, while police numbers are actually up, by 1.4 per cent year-on-year (year ending December 2014).

This seems discouraging. But it may actually mean these stats are the most reliable in years.

The problem with police numbers is the people that produce them are the same people who are judged by them. As The Wire showed, there is a crippling conflict of interest at the heart of the process.

This Office of National Statistics survey, of around 50,000 households, has been collected yearly for more than a decade – and has become the more reliable measure of crime.

So we shouldn’t necessarily be concerned that police reported crime is up for the first time in more than a decade.

Their data may just be becoming more accurate.

We can assess this by comparing the two sets of numbers. If there is a big difference between the number of crimes the ONS is reporting and the police are recording, that may indicate the police are producing ‘soft’ stats.

It appears that may have happened throughout the 1980s and early 1990s. The ONS’ survey consistently reported more than three times as many crimes as the police. But after a reporting change in the late 1990s, the difference came down.

By the early 2000s the survey was only reporting twice as many crimes as the police. In recent years that ratio began to creep back up towards a three-fold difference. But, thanks to the fall in the ONS’s measure, the difference has fallen back down to the level of the early 2000s.

Simon Jenkins, the one-time crime reporter and long-time national journalist, recently argued that “police statistics have been a conspiracy against truth for decades…The only crime figures that should count are those of the BCS.”

That may be so. But comparing the two sets of numbers can give us an indication of how unrealistic the police numbers can be.

While big falls in crime make good headlines, we should be more concerned with the legitimacy of the numbers we are fed. If today’s figures are an indication of better stats, they should be welcomed.