Last month, Anthony Reuben came to the end of an experimental 18 month contract at the BBC. His job had been Head of Statistics, and included training and advising BBC reporters on how to understand and present numerical issues. The end of his tenure was commemorated with a nice little profile at the Online Journalism blog.
Many stories that reporters get, he notes, are ‘big number’ stories which appeared to be striking but require the journalist to scrutinise further to establish whether the numbers really were striking when placed in context.
It is rather a pity Reuben didn’t stay in post for just a few weeks longer, then we might have been spared the dog’s dinner of a story which featured prominently on most BBC news broadcasts yesterday.
The headline, duly replicated in most newspapers, is captured here. “School sex crime reports in UK top 5,500 in three years.” As the broadcasts filled out the details, it was described as “a national emergency.”
The story originated on the radio documentary slot 5Live Investigates. Their reporter submitted Freedom of Information requests to every police force in the UK. The data have not been published, but here is the full methodology and results as explained by reporter Claire Savage on the show.
“We asked every police force in the UK how many sex offences were reported on school premises over the last three years.
“[They included] a wide variety of crimes from voyeurism and exposure to mainly sexual assaults. In the last three years there have been more than five and a half thousand alleged sex offences reported as crimes and more than 600 of those were rapes.”
“Although the figures may include some historical allegations of abuse by adults, which are only now being reported, we know that more than a thousand offences reported to the police in the last three years were carried out by children on children, so called peer-on-peer abuse, but because not all police forces could give us the ages of perpetrators, the actual figures for peer-on-peer abuse may actually be higher.”
“Many involved sexual assaults on children under 13 by other kids and some of the victims and alleged perpetrators were as young as five.”
Now, lets attempt to understand what this Dramatic Big Number actually means. There are slightly under 25,000 schools in England and slightly over 5,000 more in the rest of the UK, so call it 30,000, accommodating around 10 million children at any given time. The BBC is telling us that over a three-year period there was one school-based sexual offence reported to police for every 5.5 schools. Another way to present that figure is that the average British school will have a sexual offence reported on its premises once every 16 or 17 years. Annually, one child in every 5,555 will be the victim of a sexual offence at school, reported to police (10m/1800).
But wait. The content of the 5Live documentary and accompanying news reports was strictly about peer-on-peer crimes, i.e. those committed by children. As the report itself makes clear, only a fifth of reports are known to be peer-on-peer, the others could be crimes committed by staff or other adults against children on school premises, either recently or at any historical point, and indeed (I would presume) sexual assaults committed against staff by other staff or pupils. (And yes, we know that there are probably more peer-on-peer crimes within the stats than this, but we are given no clue as to how many more there might be.) All we do know for sure from the statistics revealed by the BBC is that the average British school will be the location for a peer-on-peer sexual offence reported to police about once every 80 years and that one child in every 27,000 will report such an attack each year. This is the basis for all those news headlines declaring “a national emergency.”
Now, before anyone begins to howl me down as a rape apologist or denialist, this categorically does not demonstrate that sexual assaults and other sexual offences in schools are incredibly rare. It does demonstrate that the statistics generated by the BBC’s FOI requests are almost entirely meaningless and worthless.
Once you get past the silly numbers, the 5Live Investigates programme made a whole series of really important and entirely valid points. Dr Carlene Firmin, whose excellent work has graced this blog several times before, patiently explained that the vast majority of peer-on-peer sexual offences will never be reported to anyone, or at most be dealt with as a school disciplinary matter. A couple of brave young people, one boy and one girl, told their own stories, both of which suggested that sexual assaults on school premises may not be treated with the seriousness they demand. I was particularly struck by ‘Jamie’ describing how his head teacher had tried to dismiss the sexual assault on him by three other boys as “rugby locker-room banter.”
All the expert contributors to the show stressed the urgent necessity for full and frank SRE / PSHE lessons that not only cover issues of personal autonomy and consent, but also reflect the technological environment of sexting, online pornography and all those other 21st century concerns. All of which I wholeheartedly agree with and endorse, as I have written many a time before.
The depressing fact however, is that none of these points were carried from the body of the documentary to the rest of the media, all of whom entirely ignored the valuable qualitative evidence contained within and instead reported the Dramatic Big Number [DBN] as fact, without explanation or context. Inevitably, the DBN will now be dug up and regurgitated every time a journalist wants a quick and easy demonstration of our depraved, sex-crazed teenagers (and particularly teenage boys.) Indeed this has already started. In her broadly careful and thoughtful Guardian piece this morning, Zoe Williams has thrown in the new DBN uncritically.
Some might argue that since sexual offences in classrooms and corridors are undoubtedly a real issue which causes immense trauma and stress to (quite literally) untold numbers of young people, it doesn’t much matter that the DBN is actually a chimera, a mythical representation of reality rather than an actual fact. If it catches people’s attention and inspires them to take the issue seriously, then surely job done?
This is tempting logic, but ultimately harmful. This kind of practise is a classic ingredient for moral panics. (“Now that’s a scientific fact. There’s no real evidence for it, but it is a scientific fact”) More subtly, using this type of DBN as a proxy for actual evidence serves to disguise any real trends that might be happening. Within the narrative of these reports, unspoken but unmissable, is the assumption that these problems are getting worse all the time. While these figures are all wrought with complexity, the best data we have is that the numbers of cautions and convictions issued to young people for sexual offences have been consistently declining over many years, despite the host of new technological opportunities to get in trouble. The headlines invite not only fear and anxiety among parents and young people alike, but also demand that we ask ourselves what it is we have been doing wrong, when perhaps we should be asking ourselves what we have been doing right, in order to do more of it.
The bottom line is that no one is well-served by the generation and propagation of statistical mythology. Good public policy is driven by good information and good evidence. One issue the 5Live Investigates programme did bring to light is how inconsistent and unreliable schools can be in their policies and procedures to prevent and, where necessary, respond to sexual offences in schools. The best place to start there would be with proper research to provide responsible, reliable and meaningful evidence as to the nature and scale of the problem.
[Correction note, 9/9/15: As ever when I correct people on their sums, I made my own error. This post originally said that the police reports mean one report per school every 50 years. It’s actually about every 80 years (16.5 x 5, rounded off)