Did Caroline Criado-Perez Deliberately Exaggerate Abuse?

Caroline Criado-Perez describes herself as a journalist and feminist campaigner. Is it possible to be both, without conflict? Because of the nature of their objective, campaigners are political animals. Sophisticated people understand why politicians use exaggeration, biased statistics and misinformation to persuade. But the excusable faults of a politician cannot be tolerated in a journalist. The role of a journalist is to present information, not misinformation. Journalists may be free to comment, but the information itself is sacrosanct. In our present era, the difficult dividing line between politics and journalism has been blurred to the point where nobody knows where it lies. And yet, we need dividing lines, to maintain standards in the presentation of information and the quality of public debate. Today I will be writing about Caroline Criado-Perez, suggesting that she crossed the line by grossly exaggerating the frequency of domestic abuse whilst writing for an established political journal, in order to generate sympathy for her campaigning goals.

But before I go further, it is necessary to be clear about how Caroline Criado-Perez attained a certain level of fame, which gives her the opportunity to write for widely-read publications. I must also comment on how I feel about her goals. Violence is wrong. I will not demean myself, or the victims of violence, by entertaining further argument on that point. Criado-Perez wants to live in a less violent society, and I want to live in a less violent society too. However, violence takes many forms, and is not the only evil in our world. Though violence is literally defined to be a physical act, we understand there can be violence done to people’s minds, as well as to their bodies. Deception and manipulation is a kind of violence to the mind. Sometimes one is necessary to forestall the other. The classic philosopher’s conundrum involves a homicidal maniac who mislaid his axe. You know where the weapon is. The murderer asks. Most philosophers see no struggle with the ethics of this situation: you send the killer in the wrong direction. But it is rare that we have such clean-cut consequences when contemplating lies over truth. Even well-intentioned lies cause harm. Democracy works in a free society because we trust people to generally do the right thing when they know the truth. If people have to be lied to, in order to get them to do the right thing, then that is an usurpation of democracy.

Democracy also relies upon the principle that people may speak without intimidation. Criado-Perez is famous for some relatively trivial social media campaigns, and the hostility they provoked. Getting middle-class Brits to sign a 2013 online petition saying there should be a woman (other than the Queen) pictured on the notes issued by the Bank of England is hardly as brave or as worthy an act as that undertaken by the giants of the historic struggle for civil liberties. Criado-Perez is no Rosa Parks, who refused to vacate her seat on a bus in Alabama in 1955, nor a Harvey Milk, who was openly gay whilst standing for election in the 1970’s. And yet, the banknote campaign provoked a storm of misogynistic abuse, directed at Criado-Perez. Such abuse undermines and coarsens our democracy, by intimidating the individual and levying a high price for the free exercise of their rights. All forms of intimidation, whether physical or verbal, are wrong. And that is why I will now criticize Caroline Criado-Perez for presenting deliberately misleading information, even though I know it is likely to upset her fans.

These are some extracts from Criado-Perez’s July 23 article for New Statesman:

“Isolated incidents”: how the laws around domestic violence are failing its victims

…Women murdered by men are often described by the media as tragic. There is a sense in that word of catastrophe, of horror, of something out of the ordinary. Something that could not have been prevented. Perhaps that word gives us a sense of comfort in the face of such brutality. This could not have been predicted, there is nothing we could have done. This is a freak accident.

Such words may comfort us, but they are dangerous, and our comfort comes at a cost of reckoning with a reality that we must face if we are serious about tackling the epidemic of domestic violence. And make no mistake: it is an epidemic. Domestic abuse is the largest cause of morbidity in women aged 19-44, more than war, cancer or motor vehicle accidents. It is an epidemic to which we are so inured that the steady reports of abuse, of beatings, of assaults, of imprisonment, of death, barely register. They are not front-page news. After all, to put it bluntly, “man kills partner”, is not news. It is the opposite of new. It is old. Tragically old.

She then goes on to argue for a change in the law. Criado-Perez builds her argument upon the frequency of violence. She is not arguing that a single, isolated incident of violence is wrong, and must motivate a change in law. Criado-Perez insists there is an epidemic of domestic violence, which implies the legal system does too little to address the problem.

Only one straightforward fact is presented as substantiation:

Domestic abuse is the largest cause of morbidity in women aged 19-44, more than war, cancer or motor vehicle accidents.

The problem with this fact is that it is wrong. It is very wrong. However much we may abhor violence, it is nowhere near the top causal factors of morbidity – sickness – suffered by women of any age group, whether in this country or elsewhere. That is not to say that this stat has not been published and repeated widely. It has taken on the status of an ‘urban legend’, being believed true because it, and variants of it, are reported all over the place. Google the words “more than war cancer or motor vehicle accidents” and you will see what I mean. The BBC once recited a variant of this pseudostatistic during their 10 o’clock news broadcast. Thankfully, that error prompted some statisticians who work for the BBC to debunk the stat once and for all. In 2009, the makers of the popular More or Less radio show thoroughly investigated this claim, and all its variants. They discovered that the misfact has been circulating since the 1990’s, and suffered increasing exaggeration over time. However, the statistical truth was plain: the claims that domestic violence are the leading cause of women’s morbidity are fundamentally untrue, no matter how the claims are interpreted. You can listen to the show here, and some of their key observations are repeated below.

…this claim mutates as it circulates…

…it is pretty common for a rogue statistic to spread and mutate like this, and the likely explanation is that people keep misquoting each other rather than going to a credible original source…

…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… the whole lot was booted up the league table from 6th place, to 1st place.

If you want to look for top causes of women’s morbidity, then unsafe sex is the top issue worldwide, because it leads to the top killer, which is HIV/AIDS. And when it comes to non-fatal morbidity, mental disorders do women most harm.

At this point, a generous reader may still disagree with me, or at least challenge my right to reach a conclusion. Criado-Perez is not guilty of deliberately misinforming the public, if she genuinely believes a stat which has been reported widely elsewhere. We cannot look inside her mind, to judge if she believed the stat herself. But there is another sense in which Criado-Perez misled readers, which goes beyond her choice of words. Her article was published online, and Criado-Perez inserted a hyperlink from the words “largest cause of morbidity in women aged 19-44”. The link takes readers to a 2002 World Health Organization report on global violence. It is at this point where I believe we can accuse Criado-Perez of demonstrable dishonesty, because the data included in that report contradicts her assertion.

The WHO report is long, considered and comprehensive, full of data tables, and avoids sensational language. In other words, it is the kind of report that few people bother to read. Crucially, whilst Criado-Perez’s sentence is a near word-for-word copy of sentences that can be found in lots of British reports and websites, it is nothing like any of the language used in the WHO report. For all the data it presents, and despite dedicating a chapter to violence between intimate partners, none of its contents support Criado-Perez’s claim that domestic violence is the leading cause of morbidity for women of a certain age range, whether worldwide or for any region. Nobody could suggest the chapter on domestic violence seeks to play down its seriousness, but there is never a claim that such violence outranks other causes of morbidity. And whilst there are many data tables, none of them are granular enough to map to Criado-Perez’s very specific claim about domestic violence and women aged 19-44.

The closest data that can be found in the WHO report indicates that Criado-Perez’s claim is a very severe exaggeration; table A.6 on page 286 shows that interpersonal violence is the 43rd highest cause of disability-adjusted life years – a measure of morbidity – for women worldwide. Domestic violence is necessarily a subset of interpersonal violence, but all interpersonal violence causes just 0.5% of the morbidity suffered by women worldwide. To put this into context, HIV/AIDs causes 6.5% of morbidity for women, unipolar depressive disorders cause 5.5%, malaria causes 3%, self-inflicted injuries cause 1.1% and war causes 0.4%. The table makes no mention of traffic accidents or cancer, begging the question of where Criado-Perez really sourced her statement. And there can be no refuge in the suggestion that the worldwide numbers do not apply to Criado-Perez’s argument, because she is presenting UK data as relevant to a change of law in the UK. The report contains no UK-specific stats about morbidity. The closest table covers morbidity of women in Europe, where interpersonal violence ranks 34th.

In short, Criado-Perez’s claim is not only false, it is contradicted by the WHO report she cited, as far as that report presents relevant data. So why cite a WHO report, if the data is not in there? Why take the risk that somebody would double-check the contents of the report, when she could have cited many other (unreliable) sources, without fear that they would contradict her? I suspect it is because Criado-Perez knew the other sources may be questioned, and because those other sources repeat the same assertion, but without supplying any raw data. As pointed out by the makers of ‘More or Less’, data is more credible if taken from the original source, rather than being reported second-hand. We all know that when stats are reported second-hand, they are also liable to be misreported and exaggerated. Criado-Perez believed the stat was right – and she wants it to be right – but she could not find an original source which supported the stat, because there is no original source which supports it. So she cut a corner, and cited a long and worthy UN report which tells us that domestic violence is serious, and we should be doing something about it. She hoped that nobody would try to join the dots from their data to her pseudo-data, or that if they tried, they would give up and blame themselves for not trying hard enough.

My first instincts were similar to those that many people might have. I wondered if I had read the UN report closely enough. I assumed that if a mistake was made, it was an honest one. So I left a comment on the New Statesman’s website, asking for the author to pinpoint the source of her claim. That was ignored, so bereft of other quick and convenient ways to contact the parties responsible, I tweeted both Criado-Perez and the New Statesman, succinctly stating that the BBC had debunked the stat and it could not be found in the cited report.

@CCriadoPerez @NewStatesman your stat on domestic violence was debunked by BBC and is not in the UN report you cited; please pinpoint source

— Eric Priezkalns (@ericpriezkalns) July 23, 2014

Once again, I was ignored. By this time, the story had passed, and was no longer visible on the front page of the New Statesman’s website. Most people who were going to read it, had already read it. The harm had been done. Perhaps, from Ms Criado-Perez’s point of view, that harm is justified by the good she believes she is doing. Perhaps she ignored my tweet because she dismissed me as a misogynistic hater, or maybe she ignored it because she did not want to explain why she cited a WHO report that does not include the single convenient pseudo-fact that she built her whole argument upon. I scratched my head, and wondered what to do.

And then I complained to the UK’s Press Complaints Commission. Criado-Perez describes herself as a journalist. She should be held to the same standards of accuracy expected from all journalists. The New Statesman is not without fault either, because they clearly did not verify the claim that was made against its supposed source. We shall have to wait and see how the Press Complaints Commission responds.

Do my actions make the world a better place? I am not sure, but I hope so. People trust the information presented to them, and clearly some people are far too trusting. We can agree a problem is serious without losing sight of the fact that other problems are even more serious. Violence is an emotive topic, but we make a mistake if we focus resources in a bid to massage our emotions, instead of using them in ways that alleviate most suffering. For example, mental health is woefully neglected by our society, and we will not adequately boost the care given to victims of mental illness if we choose to skew data in order to misrepresent domestic violence as a greater cause of pain. Furthermore, as I wrote in a comment to the New Statesman, I believe that grossly exaggerating violence has a corrosive effect on all of us. It cheapens the tragedy suffered by real victims, whilst encouraging a more hostile, paranoid and divisive society. To make the world a better place, we must start by seeing things as they really are, and that means we must dismiss the politicking hyperbole of self-appointed campaigners.