Crimes recorded by the police have jumped 10% overall in the 12 months to March 2017 – the largest annual rise in a decade, according to new data recently released by the Office for National Statistics (ONS).
But in the same report, the ONS reported that the latest Crime Survey for England and Wales showed a 7% decrease in crime compared to the prior year.
How can both be true? Is there really a significant increase in crime in England and Wales, or has crime actually fallen?
The new ONS figures do show some alarming rises: violent crimes surged 18% in the 12 months to March 2017. The homicide rate jumped 26%, robberies were up by 16%, and sexual assaults by 15%. Offenders were more likely to use weapons, offences with knives or other sharp instruments climbed 20% in the year, and those involving firearms increased by 23%. Police recorded property crime rates also increased. Theft offences rose 7% over the year, while public order offences increased by a dramatic 39%.
Ministers should be concerned that the numbers suggest the country may be verging on a violent crime wave. Still, there is some evidence to the contrary, suggesting that crime should not be such a pressing concern.
The two methods of studying crime statistics – a crime survey and police recorded crime – simply measure it in different ways. And they count different types of offences.
Police recorded crimes only include offences that come to the attention of police officers and are entered as official statistics. In contrast, the Crime Survey numbers are generated after face-to-face questioning of as many as 35,000 households.
The Crime Survey results include crimes not reported to police that would not appear in the number of police recorded crimes. But the survey does not capture a lot of serious offences that are counted in police recorded crimes, such as homicide, weapons attacks, and sexual assault. Officials admit that the face-to-face method means people are sometimes not forthcoming when talking about private crimes such as sexual assault. The survey method also does not cover victimless crimes such as drug possession.
In addition, unlike the police reported statistics, the Crime Survey does not count crimes occurring in communal settings, such as college dorms, assisted care facilities, or prisons.
John Flatley, a statistician with the ONS acknowledges the seeming discrepancy. In the recent report, he attributes the rise in reported crime to “ongoing improvements to recording practices”. But he also concedes that there were actual increases in crimes in certain categories.
So, which is the better method for judging changes in crime rates over time? For most low-level property crimes, which are often not reported to police, the Crime Survey is the better option. But for violent and serious crimes, the Crime Survey does not cover many of them and police recorded crime data is more effective.
This means that for those most concerned about the resurgence of violent crime, the recent substantial increases in violent offending recorded by the police is alarming.
Opposition politicians have pointed out that while reported crime was rising, substantial cuts were made to policing resources nationwide. In the Conservatives’ bid to reduce spending, the number of police officers in England and Wales has declined in recent years. Overall, police personnel are down 14% since 2010.
Despite such criticism, there is no scientific evidence that can show these reductions actually caused any increase in crime. There are just too many other factors that might explain fluctuations in crime rates, such as unemployment, the availability of social services, and the level of drug use.
The substantial increases in crime from the ONS certainly achieve bigger headlines. Perhaps the more important observation is that the results are already providing political fodder to various sides in the debate about the state of crime and policing resources. Unfortunately, the divergence in statistics means politicians can simply pick the source or statistic that suits their interests.
This article was updated on July 28 to correct the figure for the number of households invited to take place in the Crime Survey for England and Wales. It is 35,000, not 500,000.