Today can we celebrate a partial victory for truth, but the battle is not over. Two weeks ago, I wrote how feminist campaigner Caroline Criado-Perez cited a made-up statistic about domestic abuse in an article for the New Statesman’s website. She stated that British women are suffering an ‘epidemic’ of violence, and that the systemic nature of this violence has been badly underreported by mainstream British media. Her conclusion was that the scale of violence necessitates a change in British law. But only one statistic was presented amidst her extravagant language. That statistic claimed that domestic violence is the leading cause of morbidity for women aged 19-44, and the wording was hyperlinked to a report by the World Health Organization. There are only two problems with her preferred stat.
1) It is totally false.
2) It is not in the World Health Organization report that Criado-Perez cited.
Anyone who went to the trouble to read the 246-page WHO report would discover that interpersonal violence is only the 43rd-highest cause of morbidity for women. According to their data tables, interpersonal violence accounts for 0.5% of the total morbidity suffered by women. In comparison, HIV/AIDS is the genuine leading cause of worldwide morbidity for women, and is responsible for 6.5% of the total morbidity suffered. Actual rates of interpersonal violence would need to be multiplied 13 times in order to become the top cause of morbidity. Criado-Perez not only tried to use a worldwide statistic to make the case for changing British law, it is not even a real statistic.
Criado-Perez just pretended to use the WHO report as source material. Her ‘fact’ is actually a widely-spread urban legend, which can easily be found in all sorts of places on the internet. A tell-tale sign comes from how Criado-Perez worded her fact: “…more than war, cancer or motor vehicle accidents.” I dare you to google those words, to see how often they are repeated in this context. But crucially, none of those websites give an accurate report of any primary research into violence. They are merely misquoting each other, having all been written by people who wanted a juicy sensational attention-grabbing headline, but were too lazy to locate genuine research. A thorough investigation by the BBC statistics program More or Less clarified where this misinformation originally came from:
…there have been multiple bouts of statistical inflation. The 1993 table [in a report by the World Bank] showed rape and domestic violence as the 6th largest cause of morbidity in women aged 15 to 44, globally. Leaving to one side the difficulties in collecting that statistic… rape was dropped from the category without making any adjustments, and the whole lot was booted up the league table from 6th place, to 1st place.
Criado-Perez and the New Statesman ignored me, when I informed them of their error, just a few hours after the article was published online. However, they could not ignore my subsequent complaint to Britain’s national Press Complaints Commission. And so, the bogus stat has now been deleted from the amended version of the article. It is a shame that they waited two weeks to correct an error that would never have been made if they did simple fact-checking, and should have been corrected within hours of publication.
However, the damage that was done has not been undone. Almost everybody who was going to read the article would have read it before the correction was made. Readers now need to specifically search the New Statesman’s website to find the article. The New Statesman made no effort to verify the facts during the days when Criado-Perez’s piece was one of four featured articles on their homepage. I will be persisting with my complaint to the Press Complaints Commission, with a view to securing an equally prominent correction, as befits such a gross inaccuracy and such a gross failure of editorial standards.
Sadly, the editors of the New Statesman have shown they have little interest in factual accuracy. They continue to mislead the public. Instead of apologizing to readers for misleading them, they merely noted:
Update, 7 August 2014: this article was amended to change the statistic referred to in the UN report.
And is the new statistic as misleading as the old one? In a sense, no. It has the advantage of being derived from the cited WHO report, instead of being dredged from some unknown part of the internet. But in another sense, the New Statesman continues to scrabble for misleading data that preserves Criado-Perez’s argument, which was that violence against women is badly understated. This is a difficulty for them, as they can find no statistical data that supports this assertion. That is why they inserted this amended ‘fact’, in place of Criado-Perez’s bogus stat:
The prevalence of domestic abuse means that 40-70 per cent of female murder victims are killed by a husband or boyfriend, according to the UN.
It is true that WHO is a branch of the UN, and that their 2002 report referenced five studies of domestic violence, as conducted in Australia, Canada, Israel, South Africa and the USA. Those five studies give rise to this example range of 40 to 70 percent. But the authors of this WHO report never generalized from those five studies, and did not pretend this is a global statistic, like the editors at the New Statesman have.
Every single murder is regrettable and shocking. However, if we play the game of cherry-picking global stats, it is far from clear why homicide rates in Australia (10% higher than found in the UK), Canada (60% higher), Israel (80% higher), South Africa (3000% higher) and the USA (380% higher) justify claims that British women are suffering an ‘epidemic’ of violence. There is a simpler, better way to improve the New Statesman’s article. They should simply insert reliable and current statistical data about domestic violence suffered by British women. That will be shocking enough. More importantly, it will be relevant, and not misleading. The article starts by describing the murder of a British woman. It ends by demanding a change to British law. Between the two, it comments on the supposed failures of British media. So why not simply present the pertinent data, instead of shopping around for 12 year old stats from five foreign countries?
I think we know the answer. Both Criado-Perez and the New Statesman confuse their ambition as political campaigners with their responsibilities as journalists. However necessary it is to change the law, the change should be motivated by fact, not fiction. And whatever responsibility they have to inform the public about important issues, they cannot forget their responsibility to present accurate information, and to present corrections of misinformation with an appropriate level of prominence. Sadly, some people really want to believe exaggerated statistics about violence. That is unfortunate for everybody, not least the real victims of violence, whose tragedies should not be trivialized by pretending atypical suffering is commonplace. Democratic law-making serves all in society: the victims, the guilty, innocents, and bystanders. Sensational lies about the prevalence of violence will never deliver sound laws that best serve our society. Criado-Perez and the New Statesman should admit their fault, and rebuild their reputations on the solid foundation of fact. That begins with an apology to readers for misleading them, and an accurate presentation of the data.