Over the past few weeks a graph has been tweeted into my timeline several times, purporting to show that “Domestic Violence Crime has #climbed 31% since April 2010.”
The tweet was originally sent by an account called “EvidenceUK” which declares ‘The purpose of this account is to factually correct the errors and lies peddled by Tory Newspapers & MPs during the 2015 General Election Campaign.’ The graph is sourced to the Crime Survey of England and Wales, and the figures on the graph are accurate. The only inaccuracy is the detail of the tweet. The first bar does not show the data from April 2010. They actually show the data from the year before.
An accurate graph for domestic violence since April 2010 would look like this. (Note, I have taken these stats from the precise same data set linked to from the original. They are the same data)
They show not a 31% rise in domestic violence incidents – but a 3.3% fall in domestic violence since the Tory / coalition government came to power.
Now as regular readers will know, I rarely miss an opportunity to have a swipe at the Tory party and the current government, but I do also care about honesty and accuracy in media and reporting. There is a widespread myth that domestic violence has been increasing significantly since the last election, and there is not a shred of evidence that it is true.
To get an accurate understanding of what is happening with rates of domestic violence in this country, take a look at the graph over the past 20 years – again, drawn from the precise same data set linked to in the tweet above.
If you look closely you can see the historical low point of 2009-10, and a slight rise to the most recent quarterly update from July this year. However the long term trend is quite clear – domestic violence rates plummeted between the mid 90s and the mid 00s, and have been bobbing along fairly consistently ever since. Yes, they took a bit of a spike in the year to March 11, but immediately reverted the year after. In fact the ONS statisticians are quite clear that there has been no statistically significant change in the domestic violence figures, year on year, in more than a decade.
Someone looked at this whole data set to produce this graph, and must have known exactly what they were doing when they cherrypicked a statistical lowpoint to draw their comparisons. This type of statistical legerdemain is a source of constant annoyance and frustration, more so when it comes from people with whom I would like to be on board. There are plenty of reasons to despise the current government and plenty of genuine reasons to condemn their track record. Mischief like this simply makes me lose faith in those sharing the information, and that helps no one.