Is violent crime going up or down?
The claim
“This new analysis confirms that the level of violent crime actually reported to police officers in police stations up and down the country is much higher than it was a decade ago.”
Chris Grayling, Shadow Home Secretary, Press Association, 9 March 2010
Cathy Newman checks it out
To see Labour and the Conservatives trade blows on crime figures, you’d be forgiven for thinking that there are lies, damn lies and home office statistics. Last month, the Tories got in a muddle when they claimed violent crime had increased under Labour. Today, they returned to the fray, wheeling out a magisterial House of Commons statistician to back them up. So are they any closer to the truth now?
The analysis
Last month the Tories earned a rebuke from the statistics watchdog when they claimed violent crime had increased under Labour.
The Conservatives were using the number of violent crimes recorded by the police. But in 2002-3, a big change to the way in which crime was tallied meant contrasting the figures before and after that date is like comparing apples and pears.
When the new counting rules were brought in, police had to record any incident where the person considered themselves to have been a victim of a violent crime, regardless of the outcome of proceedings. This meant crimes that may not previously have stood up to the police definition of a violent crime started to show up in the figures.
The Home Office estimates that the new rules accounted for an apparent 23 per cent increase in violent crime.
So the Tories asked the House of Commons library statistician to adjust police recorded crime figures for the effect of the 23 per cent increase. In 1998-9, there were 502,778 violent offences recorded by the police in England and Wales. Multiply this by 23 per cent, and you get 618,417 offences. Compare the inflated 1998 figures to today’s, and it would seem that crime has increased by 44 per cent since 1998.
The Tories refused to publish the document in full.
But the government hit back, saying that violent crime in England and Wales had in fact fallen by 41 per cent (half a million fewer victims) since 1997. How so?
As it has done in the past, the government is using a different series of crime figures – the British Crime Survey.
The BCS has been around for nearly 30 years, and asks people whether they have been a victim of crime in the past year. So it’s based on what people say about crime when they’re surveyed, as opposed to incidents that are actually reported to the police. We’ve looked at the differences between this and police recorded crime several times, see here or here for more explanation.
The Home Office says the BCS better indicates long-term trends because it isn’t affected by changing levels of reporting to the police.
This doesn’t mean it’s perfect. The Conservatives pointed out that there are a number of exclusions from the BCS, including murder and (until the as-yet-unpublished 2009 research) crimes experienced by under-16s.
So who’s right – Labour or the Conservatives?
Cathy Newman’s verdict
Because the Tories are quoting from recorded crime figures and Labour are using the British Crime Survey, they’re both coming up with different snapshots that suit their own political agendas. There’s no doubt the Conservatives gaffed originally when they failed to take into consideration the new counting rules, so today’s figures are a definite improvement. But the statistics watchdog boss Sir Michael Scholar hasn’t finished with the Tories yet. He warned the shadow home secretary yesterday that the “selective quotation” of one series of crime statistics without the other “could prove misleading”. But that’s unlikely to stop either party spinning the figures as polling day approaches.
Update 1: Home Secretary Alan Johnson told Cathy Newman: “As the UK Statistics Authority has made clear, the British Crime Survey is the most reliable way by which crime is recorded. That clearly shows a 36 per cent reduction in crime and 41 per cent reduction in violent crime since 1997.
“Chris Grayling has been trying to mislead the public about crime stats but has been caught and reprimanded. It is vital that the Conservatives are not allowed to continue to mislead the public about this.”
Update 2: This House of Commons briefing paper, Trends in crime since 1997, was also released on 9 March. Was this where the Tories got their figures from? Although the paper contains a graph (last page) and some information on adjusted reported crime rates, it certainly didn’t seem to have enough to back up their claims.
However, a Conservative spokesperson said this was not the mystery document. The party were given data separately on crime rates since 1998 rather than 1997, as 1997 data contained six-months worth of crime rates under the last Conservative government.
There are 7 comments on this post
It’s a depressing yaa boo because neither side is actually threatening to change much and since in a couple of years their economic policies will be equally austere why should anyone vote other than to make a statement against the fascists and knock their percentage.
The Tories here just quoting numbers they know to be wrong or less accurate for advantage. No real ideas or policies and indeed their attitude to high crime groups like drug addicts is likely to make it worse.
Like or Dislike:
0 
0
The BCS does cover under 16s http://www.homeoffice.gov.uk/rds/bcs1.html
and since the whole point is it’s based on interviews with victims of crime, it’s never likely to include murders.
Pete
Like or Dislike:
0 
0
Thanks for the comment. The BCS does now survey under-16s, but only since 2009, and therefore not in any published results. We’ve now updated the text to clarify this.
Like or Dislike:
0 
0
So what is your verdict, fact or fiction?
Like or Dislike:
0 
0
The head of the Statistics Authority said in his letter to Grayling that, “..a more balanced commentary on national trends in violent crime would, in the view of the Authority, also make reference to estimates given in the British Crime Survey, which in our view provide a more reliable measure of the national trend over time.”
Like or Dislike:
0 
0
If the Government believes that the British Crime Survey statistics are more reliable than the police’s own figures one does have to wonder why the Government requires the police to take so much time recording figures that even the Government says are unreliable.
Also, has there been a change in which set of figures the Government chooses to use? I could be wrong, but I seem to recall that years ago the Government relied on police figures to say the BCS was wrong, now they seem to be doing the reverse.
Like or Dislike:
0 
0
Kathy Newman says that the BCS do consult under 16′s. As far as I can make out in January 2009 the BCS PROPOSED to extend the consulting age to under 16′s.
There is a difference in proposing and actually doing, can anyone elaborate. Are the BCS NOW at this moment consulting under 16′s.
In any event any figures they give prior to January 2009 should still be relevant surely.
In my view the BCS figures are unreliable as they are capped at 5 crimes per person per year, which is unrealistic when it concerns domestic violence.
Also,they only cover the victims of crime, so do not cover, commmercial/business crime, crimes committed in universities,and hostels and care homes, drug crimes,murder/manslaughter, and car crimes.
Slashing of car tyres, damaging vehicles, and throwing bricks at victims doors can be pretty violent, as far as they are concerned.
Come on Cathy I know you work for channel 4, and therefore most probably are biased, but you would do your readers more good if you were to rise above this and give the true facts.
Like or Dislike:
0 
0