Are police solving fewer crimes?

"Cops solve 53,000 fewer offences"

"The drop was the first time in a DECADE the detection rate has fallen. Just three in ten crimes were solved in 2011 — meaning 2.8 MILLION went unpunished"

The Sun, 1 June 2012

Today the Sun claimed that the police were solving fewer crimes and that, in 2010, they solved just three out of ten crimes. This was supposedly the first time in a decade that the detection rate has fallen.

Analysis

The Sun cites the Ministry of Justice as the source of their figures, however figures produced by the Home Office actually provide the figures upon which the article is based. Oddly, these were published midway through last year, so it isn't yet clear why the Sun reported the figure today.

The relevant figures from the Home Office show the number of detected crimes which includes data provided by the 43 territorial police forces in England and Wales and the British Transport Police.

Detected crimes are those crimes that have been 'cleared up' by police or those crimes that have been solved. There are two: 'sanction' and 'non-sanction' detections. 

According to the Home Office, Sanction detections are detections where offences are resolved through a formal sanction - including being charged or receiving a caution. Not all sanction detections lead to a conviction. In cases where a defendant is charged the Crown Prosecution Service (CPS) may not decide to proceed with the case and the defendant can always been found not guilty.

Non-sanction detections are where offences are counted as cleared but where no further action is taken. Since April 2007 there are only two reasons for a non-sanction detection: the offender dies before proceedings are completed or the CPS decides not to prosecute even when they are satisfied there is enough evidence to provide a realistic chance of conviction.

Since 2008/9 non-sanction detection figures also include figures for Youth Restorative Disposals (YRDs). YRDs are being piloted across eight police forces and allow police officers to deal with low level crime among 10 — 17 year olds where it is not in the public interest to prosecute. They usually involve some form of restorative justice, such as a meeting between the offender and victim.

The change in 2007 led to a large drop in the number of non-sanction detections and this combined with the fact YRDs are being piloted in a few areas should be remembered when using non-sanction detection figures.

The table below shows the number of detections and the sanction detection rate in 2009/10 and 2010/11.

The table shows that, between 2009/10 and 2010/11, the number of sanction detections fell by 53,328 - which was reported accurately in the Sun.

However, over the same period, the number of offences fell by 188,275. When we actually look at the sanction detection rate it is the same in both years at 27.8 per cent (although precise calculation shows the rate has fallen marginally).

So the Sun is correct to report a fall in detected crimes and technically correct to report a fall in the detection rate - although this fall is negligible and doesn't even show up on the rounded Home Office figures.

The data also supports the claim that three in 10 crimes go unresolved - and subtracting detected offences from total offences gives just under three million unresolved offences in 2010/11.

But is this the first time in a decade the dection rate has fallen? The following graphs show the sanction detection rate and the total detection rate (including non-sanction detections) since 2002/3:

The sanction detection rate and the total detection rate have risen overall since 2002, although both have somewhat flatlined for the last five years. The sanction detection rate actually fell in 2003/4, though the total detections ( including non-sanction detections) did not.

But is this a useful measure for judging police performance?

Not necessarily - the detection rate is not the best way to measure police investigative performance. The Home Office says:

"Detection rates are not a direct measure of police investigative performance and need to be interpreted with care. For example, some of the offences with the highest detection rates are the offences most influenced, in terms of their recorded numbers, by proactive policing to apprehend offenders (for example, drug offences and many of the offences in the 'other offences' category)."

So detection rates can actually be affected by different policing tactics for different offences. In addition, a crime may be considered 'solved' even when a detection is not made, for example when the police are satisfied they know who the offender is but where a victim is unwilling to cooperate any further.

Conclusion

The Sun are correct to say that the number of detected crimes has fallen by 53,000, and that detections now account for around three in 10 offences. However, while the detection rate has fallen, this fall is negligible, and the rate is essentially unchanged since last year.

However whatever drop might have taken place is not necessarily due to poor police performance. The Home Office has itself stated that the detection rate itself should not be used as a measure of police investigative performance.

However the evidence suggests that this is not the first time in a decade detection rates have fallen - whether sanction detections alone or counting overall detections. Nevertheless the reality of the figures is that detection rates are unchanged and have been broadly so for the last five years.

So there is perhaps less of a headline here than the Sun's reporting would suggest.