A confusing picture

Coverage of last week’s (25 January 2018) crime figures (which report on crime for the year ending in September 2017) was generally confusing in a manner which we have come to expect with different commentators claiming both that crime was up and that it was falling.
Crime statistics (produced by the Office for National Statistics) rely on two main sources: offences recorded by the police and the results from the Crime Survey for England and Wales (CSEW).
Both sources have limitations. Police records are subject to regular, and often quite significant, ways in which particular crimes are committed. There are also very large numbers of minor offences which are either never ported to or recorded by the police.
On the other hand, CSEW does not ask respondents about their experience of very serious crimes and only people in mainstream residential accommodation are surveyed. Therefore CSEW records no business crime and also excludes the experiences of those living in institutions without a permanent address, many of whom we know to be extremely likely to be the victims of crime.
So, what is happening to crime?
Well, the most careful analysis suggests that serious violent offences (and possibly thefts of car) are going up while crime overall and even some forms of fraud are probably going down.

The detail

Overall crime trends

Crimes covered by the CSEW increased steadily from 1981, when the survey began, before peaking in 1995. Since then, there have been substantial falls across most forms of crime covered by the survey. The main drivers have been declines in high-volume property crimes such as vehicle-related theft and criminal damage. While the latest CSEW estimate of 10.6 million crimes against the household population seems like a big number, in the context of the overall population, becoming the victim of crime is not a common experience for most people. The survey shows that the large majority of adults (8 in 10) were not a victim of any of the crimes asked about in the survey in the previous 12 months. Including new estimates of fraud and computer misuse, around 2 in 10 adults were victims of one of the crimes asked about in the latest survey year.

Violent crime

While recent estimates from the CSEW show that most of the higher volume crimes either fell or were at a similar level to the previous year, the police recorded continuing rises in a number of higher-harm violent offences that are not well-measured by the survey. This was most evident in the relatively low volume offences such as knife crime (up 21% to a total of 37,443 recorded offences) and gun crime (up 20% to 6,694 recorded offences). The occurrence of these offences tends to be disproportionately concentrated in London and other metropolitan areas. While it is possible that improved recording and more proactive policing has contributed to this rise, it is the ONS judgement that there have also been genuine increases.
The total number of homicides recorded by the police fell by 1%. However, recent trends have been affected by the recording of incidents where there were multiple victims such as the terrorist attacks in London and Manchester and events at Hillsborough in 1989. If these cases are excluded the latest figures show that there were 57 more homicides than the previous year, a 10% rise up to a total of 650. It is too early to judge whether this represents a change to the long-term downward trend.
While, for many types of offence, police recorded crime figures are not thought to provide a reliable measure of trends in crime, certain crimes are thought to be less affected by changes in policing activity and recording practice. The police recorded rises in a number of these categories including vehicle-related theft offences (up 18%) and burglary (up 8%). These are crime types that generally have high levels of reporting to the police by victims.
The graphic below shows police figures for the more serious crimes showing comparisons with the previous year and five and ten years ago:

However, although the most serious violent offences seem to be increasing, overall it seems that we are less likely to be a victim of violence.

There were an estimated 1.2 million incidents of violence experienced by adults aged 16 and over in the latest CSEW survey year ending September 2017; no change from the previous survey year (the apparent 11% decrease was not statistically significant).
This fairly flat trend continues that seen in recent years, with no significant year-on-year change since the survey year ending March 2014. However, the cumulative effect of this downward trend has seen a statistically significant decrease of 29% in the latest survey year compared with the year ending March 2013.

The longer-term reductions in violent crime, as shown by the CSEW, are also reflected in the findings of research conducted by the Violence and Society Research Group at Cardiff University. Findings from their annual survey, covering a sample of hospital emergency departments and walk-in centres in England and Wales, show that serious violence-related attendances in 2016 showed a 10% fall compared with 2015, continuing a long-term downward trend.
Around 2 in every 100 adults were a victim of CSEW violent crime in the latest survey year, compared with around 3 in 100 adults in the survey year ending March 2007 and 5 in 100 adults in 1995 (the peak year).

Conclusion

The complexity and nuance of the statistical picture does not align well with mainstream media coverage (especially since headline writers — see above — inevitably focus on bad news).
The true picture, however, seems to be that we are still in a period of long decline in the amount of crime committed but that there are growing concerns about more serious violent crimes including knife crime. We shall have to wait another three months to get an idea whether these are worrying new trends or blips.
[Thanks to Nick Hillier for use of the header image.]

Criminal Justice posts are sponsored by Get the Data

Measuring social impact

Our cutting-edge approach to measurement and evaluation is underpinned by robust methods, rigorous analyses, and cost-effective data collection.

Proving Social Impact

Get the Data provides Social Impact Analytics to enable organisations to demonstrate their impact on society.