London 2012: What is Boris Johnson's record on crime?
18th Apr 2012
"Crime in London has been cut by 10.8%"
Boris Johnson campaign, Official Website
"Boris promised 20% cut in crime last election still standing despite crime only down 5.8%"
Ken Livingstone campaign, 11 April 2012
"Under Boris, crime has fallen 10.6%, robberies are down 16.7% and murders have decreased by 24.4%"
Boris Johnson campaign, 11 April 2012
"Knife crime, rape, burglary and robbery all up under Boris Johnson"
Ken Livingstone campaign, 11 April 2012
"overall, knife crime has declined in the capital"
Boris Johnson, London.gov.uk
During an election campaign, accuracy is paramount. Without it, people can form opinions and make important decisions based on bad information.
With two weeks remaining before London heads to the polls to elect its next Mayor, there is a risk that conflicting claims from the two main candidates on crime - one of the key policy battlegrounds - could leave voters confused.
So why are Boris and Ken's claims so different, and are they accurate?
Nationally, crime is measured using either 'police recorded crime' or the 'British Crime Survey' (BCS). However BCS statistics for London are less specific than recorded crime figures, and the Metropolitan Police only regularly publish recorded crime statistics.
For this reason we will analyse the candidates' claims based on recorded crime, although it should be remembered that these do not capture all crimes that are actually committed. You can find out more about the most recent survey of London crime here, and about the BCS from the Home Office crime statistics User Guide.
Police crime statistics in London are easily obtainable from the Metropolitan Police website, although until recently these were only provided by financial year (April to March the following year). The Met Police do not regularly publish more detailed statistics, with the rare exception of monthly knife crime statistics dating back to 2008.
Boris Johnson became Mayor of London in May 2008, so the financial year April 2008 to March 2009 almost exactly captures his first year in office. We will use the financial years 2008-09 onwards as a proxy for his time in office, with those before representing Ken Livingstone's mayoralty.
In addition, Full Fact uses no crime data prior to 2002. That was when the Metropolitan Police adopted the National Crime Recording Standard which changed the way crime in London was recorded. Trends prior to 2002 are not reliably comparable to those since.
Finally, there are two police forces in London, the Metropolitan Police and the much smaller City of London Police. References to 'London' only concern the Metropolitan Police and their area.
How Boris and Ken measure crime
Boris Johnson has repeatedly asserted that crime in London is down by around 10 to 11 per cent on his watch. These figures change slightly over time as more up-to-date statistics become available to the candidates.
The claim that still displays on the BackBoris website puts the drop at 10.8 per cent. This comes from a Mayor's Office for Policing and Crime (MOPC) monthly report from March this year, and can be seen in this table from that document:
The statistics provide data up to January 2012, so the most recent figure will be slightly different from the 10.8 (as evidenced by altering claims from the Boris campaign). However the table provides evidence of Boris's method of reporting crime.
Boris's method is, in essence, to count the total number of offences that were recorded during his time as Mayor - May 2008 up to the present day. He then compares this to the total during the same number of months under Ken. In this case, all the crimes recorded from August 2004 to April 2008. This shows a drop of nearly 11 per cent in all notifiable offences.
Ken Livingstone's campaign say crime is down by a much lower amount - approximately 6 per cent.
A thorough look through the Met Police's crime statistics reveals the source of this calculation. According to the latest stats, in Ken's last year as Mayor (2007-08) there were 862,032 offences recorded by the police. By 2011-12, this figure had fallen to 812,566. This represents a fall in crime of 5.7 per cent.
The two candiates' claims are therefore based on different analyses of exactly the same statistics from the Met Police:
Exactly the same difference in method causes both candidates to claim that robbery is going both up and down. The graph below shows how this can make a severe difference to how the crime is reported:
So what can we say about these methods? Without looking further than the two trends shown above, it is obvious that Boris Johnson's method is problematic. By taking both his and Ken's entire time in office and comparing 'lump sum' number of crimes committed in each, he fails to report the actual trends during his time in office.
Those trends clearly show that robbery offences have grown in every year since 2008, and are higher today whether you compare it to Ken's last year or at the end of Boris's first year. There is however a secondary question of how great the Mayor's influence over crime rates really is.
In addition, crime continues to fall as it has done consistently since at least 2002. The rate of decline has decreased since 2008. Again, we can't necessarily attribute this to the current Mayor's policies.
So while Boris would be technically correct to state that there have been fewer crimes committed under him than under his predecessor, this actually says nothing about whether those crimes have gone up or down during his own time in office.
Ken Livingstone's method needs to be carefully understood as well. It captures trends during Boris' time in office, but is less good at showing longer-term crime volumes.
You can see all the data we have used in our research here and below:
What this data doesn't show are figures for knife crime, which are measured as subsets of the existing offence categories (for instance, robberies can involve knives or sharp instruments). However as mentioned earlier the Met Police do publish separate figures to account for this.
The MOPC report from which Boris Johnson draws his other figures for crime provides a summary of knife crime offences since he took office:
There is a very slight trend upwards since the Mayor took office - something the MOPC attribute to rises in personal robbery over the last 12 months in particular. So it isn't clear what Boris means when he claims that knife crime "overall" is down.
Should we even care about the numbers?
While both candidates have been prolific in their use of crime numbers, it is worth taking a step back and asking whether any of this debate is at all useful. Only looking at crime numbers in London misses out two crucial factors:
1. The crime rate - i.e. how many crimes occur as a proportion of the population. A growing population can mean more potential for crime.
2. Whether London is any different to the national trend. If all of England and Wales is experiencing similar shifts in crime, what difference does a Mayor really make?
The crime rate
Crime rates are measured as standard practice by the Home Office in their 'Crime in England and Wales' series of releases, and are also broken down by police force. Measuring the rate requires estimates of the population in each force, which the Office for National Statistics (ONS) duly provides.
Using these estimates allows us to more usefully examine how crime rates have changed under Boris Johnson.
Looking at overall crime, the number of offences has of course fallen by around 6 per cent on Boris's watch - from an 'inherited' count of 862,032 offences to 812,566 in 2011/12.
But because London's population is increasing, the crime rate has fallen slightly more impressively. In 2007/08, there were 114.4 crimes per 1,000 people in London. By 2011/12 this had fallen to 104.7 crimes per 1,000.
The crime rate in London has therefore fallen by 8 per cent - more than the 6 per cent fall seen in the raw numbers. The graph below shows how the crime numbers and rates have fallen since 2002:
London versus England and Wales
However the whole debate over how the Mayor has affected crime in London seems pointless without a perspective on what has been happening in the rest of the country and how it compares to London's crime levels. If exactly the same crime trends are occurring across England and Wales, has the Mayor really made any difference at all?
To help answer this, Full Fact brought together figures from Crime in England and Wales data, both recent and dating back to 2002, as well as population estimates for England and Wales from the ONS and those for the Metropolitan Police area so that the national crime rate can be compared with the London crime rate.
London's crime rate is significantly higher than the national average, although this has consistently been the case since 2002. London's current crime rate is approximately 105 crimes per 1,000 people, compared to a national rate of 76 (in 2010/11 - so slightly dated by comparison).
What we really want to know is how quickly the London rate has fallen since 2002 compared to the national drop over the same period. By again using an index - setting 2002 at '100' and charting the relative decline since - we can see exactly how much the rates have fallen by in the two areas:
Created with flickr slideshow.
The graph for total offences above shows that the crime rate in London fell more sharply than the national average from 2002 to 2008 - although London performed only very slightly better. Since 2008, London's fall in crime has slowed while the national picture (as far as data to 2010-11 can show) has recently seen steeper falls in crime rates.
Not a great deal can be read into the slight differences between how steeply the two rates are falling. What the picture does tell us is that, for overall crime, London isn't actually all that different to the rest of the country.
However, looking at robbery shows a markedly different trend, with the rate in London rising year-on-year while the national rate holds steady. This only presents more concerns about the incumbent Mayor's claims about robbery (although robbery rates tend to be more volatile than overall crime, owing to the smaller number of incidents recorded).
The source of confusion surrounding the claims made by Boris Johnson and Ken Livingstone boils down to two key factors: their methods of measuring crime, and the timeframes they are using when quoting figures.
Boris counts up all the crimes of different types that happened under Ken Livingstone's second term and compares them to the sum total during his premiership. Meanwhile Ken takes the crime figures as they were when he left office and compares them to what they are now.
There is little doubt that Boris's method of measuring crime in this way is seriously problematic and masks crime trends during his own time in office. In particular, robberies and robbery rates, as measured by police recorded crime, are increasing in London faster than they are at a national level - which is broadly flat.
Knife crime offences show a slight upward trend in London, so it is unclear what Boris means when he says knife crime has declined in the capital - especially given that statistics before 2008 for this crime are difficult to come by.
However both campaigns are enormously selective in the crime measures they choose to talk about. Boris focuses on crime overall, murders, and specific crime indicators on London's public transport. Ken focuses on the few things that have actually increased since 2008 - namely robbery, rape, knife crime and burglary (although the burglary rate is broadly the same).
This does very little to give Londoners a clear picture of what is happening to crime in the capital.
But as Full Fact has also shown, the whole debate is potentially a nonsense without also looking at what crime is doing across England and Wales. These statistics betray very little difference overall between crime in England and Wales and crime in the capital.
So in spite of all the claims about crime in London, a better debate is needed if we are to understand whether the Mayor has actually made any substantive difference to the crime figures. At first glance, there seems to be little supporting evidence.