The shifting sands of recorded crime

Blog post

The shifting sands of recorded crime

Last month the Evening Standard newspaper reported “Violent crime in London soars by 22 per cent”. Dig a little deeper however, and this seems to have had little if anything to do with Londoners suddenly attacking each other in greater numbers. It also seems to be indicative of changes to crime recording practices nationally that increasingly call into question the use of police recorded crime as a reliable basis for assessing either crime trends or police performance.

First, those Metropolitan Police (MPS) figures. Last October, senior officers were grilled by Mayor Boris Johnson and his Deputy Stephen Greenhalgh at a Mayor’s Office for Policing and Crime Challenge Panel on police performance. Assistant Commissioner Helen King stated that the proportion of violent incidents reported by the public by telephone that are subsequently recorded as crimes has increased from 40 per cent to 75 per cent over the last two years. So much for violent crime soaring much of the explanation seems to be that the MPS has taken a stronger line on recording practice.

At a national level, the recent HMIC Crime Recording: Making the Victim Count report includes an intriguing graph (on p.55) from the Office for National Statistics (ONS). This shows the proportion of crimes that respondents to the Crime Survey for England and Wales (CSEW) say they reported to the police (scaled up from the sample to the national level) that actually were recorded by the police.

In the 1980s and 1990s it sat somewhere between 50 and 60 per cent. Following the introduction of the National Crime Recording Standard (NCRS) it then peaked at just over 90 per cent in 2003/04, before gradually tailing off to around 70 per cent in 2012/13, presumably as the initial NCRS effect reduced and the focus on data quality similarly diminished. That was also the height of the performance management era in policing. While the CSEW shows many crime types falling, this suggests that a progressive deterioration in recording practices has driven police recorded crime down even faster, at least until recently.

In 2013/14 the proportion of crimes recorded nationally jumped from 70 per cent to just above 80 per cent in a single year, apparently reflecting the “soaring” violence in London. The explanation seems simple: the microscope was turned on data integrity by the Public Administration Select Committee, the Office for National Statistics and HMIC, and forces responded in a clear example of “what gets measured gets done”.

Back to the HMIC Crime Recording report, we find that in the year from November 2012 to October 2013 police forces nationally only recorded 81 crimes out of every 100 that they should have. HMIC described this as “indefensible”. Although the figures are not statistically reliable at a force level, it is nevertheless clear that some forces seem to do a very bad job of crime recording, while others seem to do a very good job indeed (note that HMIC found no great concerns about how crimes were classified, rather the problem was whether they were recorded at all).

So there is a very mixed national picture, and some forces’ crime data are “better” than others. I am told that at least one force that did very badly had empowered all of its officers to record crimes as part of a move towards recognising their professionalism and implementing the Code of Ethics. In that case, the best of intentions (demonstrating empowerment and trust) seem to have been applied in the wrong context, where in fact a degree of centralisation and specialisation is necessary to get the details right first time.

Why does all of this matter?

First, because it implies that (all else being equal) the service provided to victims has generally deteriorated over time as proportionately fewer have had their crimes recorded (and indeed investigated).

Second, because it points to the partial understanding at least some forces must have of the underlying demand for their services, which inevitably plays into the debate about resources and cuts.

Third, because, as was the case with the introduction of NCRS, longer-term trends are complicated by changing recording practices, making it difficult to determine underlying crime trends at force and more local levels (where the CSEW sample is too small to provide a reliable picture).

Fourth, because in an era of austerity it highlights the importance of continuing to invest in the unglamorous work of auditing, quality control, training and business support functions.

And, finally, fifth, because HMIC is going to be inspecting forces on how effective they are at cutting crime, and it is far from clear how, in doing so, they will be able to take into account the shifting sands of crime recording practices.

Gavin Hales,Deputy Director, The Police Foundation

This blog was originally published on 12 February 2015 by Police Oracle.