I have nothing but admiration for the good folks at Full Fact and the time and effort they put into monitoring the accuracy of factual claims as they appear in the media, but even they sometimes miss a trick or two and their analysis of the media coverage of an NSPCC press release on the scale of child sex offences in England and Wales is prime example of the kind of errors that can easily creep in when dealing with complex issues and an equally complex set of statistics.
So, and this a very unusual occurence, it becomes necessary for me to fisk one of their analyses.
To begin, I’ll include Full Fact’s preamble, which – as ever – does a very good job of clarifying the issue were dealing with:
Is a child sexually attacked every 20 minutes?
Following a series of Freedom of Information requests by the NSPCC, several newspapers reported that a child was sexually assaulted every 20 minutes in 2011. Did they report the figures accurately?
“Child sex offence every 20 minutes” – The Sun, 4 April 2012
This week several news outlets ran stories highlighting the high rate of child sex offences in 2011. The most startling claim stated that a child was sexually assaulted every 20 minutes last year.
But are the claims made in many newspapers an accurate reflection of child sex attacks in Britain today?
That’s a very good question, especially as the NSPCC’s press release and statistics have reported in an entirely uncritical manner by the media despite the charity having made full use of the much the same style of reporting that has previously been used by rape campaigners and which has attracted a considerable amount of criticism for its use of statistics.
So, with that in mind, let’s start to look at what Full Fact made of these claims.
The press release from the NSPCC defines a child sex offence as covering various sex crimes:
“including rape, incest and abuse of children through prostitution and pornography.”
Straight away we have an issue that Full Fact have missed.
The media has largely reported this story in terms of ‘child sex attacks’ and sexual assaults and not all child sex offences in law can be reasonably categorised as either and assault or an attack, most notably the offence of ‘sexual activity involving a child under 16’, which ws introduced by the Sexual Offences Act 2003 replacing the previous offence of ‘unlawful sexual intercourse’. These are offence in which the victim will have freel consented to the sexual activity albeit while at an age at which the law does not permit them them to give such consent and, as such, these offence cannot reasonably referred to as either attacks or assaults as this implies that the offence were non-consensual.
The figure used in most of the news headlines is that every 20 minutes a child is sexually assaulted in Britain. This particular statistic was not, however, included in the NSPCC press release. The NSPCC said:
“Figures obtained by the NSPCC reveal sixty child sex offences a day”
This calculates at roughly the rate claimed by the news outlets. To check this figure further, we asked the NSPCC to provide us with the raw data they obtained from the various police services.
They provided us with this table of data retrieved from all 43 police forces in England and Wales. This confirms the total number of offences for under 18 year-olds across all the forces was 23,097, although this is subject to different measurements used by each force.
Dividing this figure by 365 gives 63.2 offences a day or one every 23 minutes.
The table supplied to Full Fact by the NSPCC does not provide a breakdown of their figures by either offence or age of the victim, only by the victim’s gender and police force area. For reasons that will become clear shortly, this is an important point of qualification when it comes to assessing the accuracy of both this data and the manner in which it has been used by the NSPCC and reported in the press.
The annual figure of 23,097 was said to account for over one third of the total number of sexual offences committed in 2011 which was 54,982. This figure can be confirmed from the Home Office Report on Crime in England and Wales in 2010/11.
This is fine as far as it goes but, as ever, the devil is in the detail.
Another issue highlighted by numerous news outlets was the significant difference between boys and girls when it comes to sexual abuse. As the Express reported:
“six times as many offences were committed against girls (19,790) than boys (3,218).”
These figures can be verified from the information sent to use by the NSPCC which shows that of the 23,097 child victims of sexual assault 3,218 were male, 19,790 were female, 88 did not have their gender reported and one did not have their age reported.
Looking at victimisation data or sexual offences from the British Crime Survey, the female to male ratio for adults is typically somewhere between 5 to 1 and 6 to 1, so these figures should really not come as any great surprise.
It is important to note however that different police forces can collect data on sexual offences differently and this can affect the quality of the statistics.
For example, the NSPCC in the press release and in the information provided to us have a figure of 426 under 18s being previous victims of a sexual offence. However, only one third of police forces were able to supply the NSPCC with this information. Forces that were unable to provide the NSPCC with this information include the Metropolitan police force – which at 3,420 had the largest recorded number of child sexual offences.
More importantly the age at which the law considers a victim to be a child varies depending on while sexual offence you’re talking about.
For the majority of sexual offence, including the most serious non-consensual offences, i.e rape and sexual assault, a child is defined as anyone under the age of 16 but for small number offences, those involving the making or possession of indecent images, abuse of children through prostitution or pornography, abuse of a position of trust and familial sexual offences, i.e. incest, the law defines a child as anyone under the age of 18.
For that reason, its always wise to consult the Crown Prosecution Service’s legal guidance on the Sexual Offences Act 2003 before delving into the statistics.
Where this becomes a matter of particular concern is when one reads the NSPCC’s actual press release and you find the following statements:
The NSPCC is calling for action to drastically reduce the number of sexual assaults on children after new figures revealed there were more than 400 offences reported to police every week last year (2010-11) with fewer than one in ten resulting in a conviction.
Of the 23,097 victims more than a fifth – 4,973 – hadn’t reached secondary school age and almost 1500 were five or under. However the majority of offences, 14,819, were reported against 11-17-year-olds.
The figures, obtained by the NSPCC from all 43 police forces in England and Wales, cover various sex crimes, including rape, incest and abuse of children through prostitution and pornography. They reveal more than one in three of all sex crimes* are committed against children and at 19,790, the number of girl victims was six times higher than boys – 3218. In 89 cases the gender was not recorded.
The statistics also reveal that 426 of the under-18s had previously been victims of sex offences but only one-third of the forces were able to supply this information.
Now we have two problems with the NSPCC’s data.
Not only have they muddied the waters by failing to acknowledge as distinction between non-consensual and consensual offences by treating the terms ‘sexual assault’ and ‘sexual offence’ as if these are legitimately interchangeable, when they’re not, but they also appear to have frame their FOI requests to these police forces in terms of offence reported where the victim was under the age of 18, using this as their definition of a child, when the majority of offence, including all the most serious offences, define a child in term of their being under the age of 16.
At this point, we need to delve into the statistics ourselves because these have, in fact, been published by the Home Office and unlike the NSPCC’s figures they do provide a breakdown by type of offence and, in many case, by the age of the victim.
For the offences in which a child is defined as being under 18, the national figures for England and Wales for 2010-11 were as follows:
Abuse of children through prostitution or pornography – 152 reported offences
Abuse of a position of trust – 146 reported offences
Familial Sexual offences – 808 reported offences (although this may include both an unknown number of offences where all parties to the offence were adults and offences that were retrospectively reported in 2010-11 but which occurred several, if not many, years previously so this figure needs to handled with a great deal of caution.)
For non-consensual sexual offence, i.e. rape and sexual assault there were 11,458 offences recorded in 2010-11 which verifiably relate to offences against children under the age of 16.
The official recorded crime statistics provide separate figures for rape offences against: –
– males aged under 13 (671 offences),
– males aged under 16 (247 offences),
– females aged under 13 (2,235 offences), and
– females aged under 16 (2,879 offences).
They also provide figures for sexual assault against children under the age of 13 for which male victims accounted for 1,125 offfences and female victims for 4,301 offences.
However, and rather unhelpfully for or purposes, the remaining figures for sexual assaults are given for males aged over 13 (1287 offences) and females (16,358 offences).
An unknown proportion of these offences will relate to victims who were under the age of 16, although the correct figure (whatever that is) will be lower that the number of sexual assaults included in the NSPCC’s figures are these are based on an incorrect definition of the offence – a sexual assault on a 16 or 17 year old, of either gender, is treated in law as offence against an adult and not as a child sex offence.
There were also 310 recorded sexual grooming offences, all of which would be child sex offences (but not assaults or attacks) and 67 offences of trafficking for sexual exploitation of which its possible that some my relate to children.
So far, we have a total of 12,066 verifiable child sex offences to which we need to add unknown proportion of the 17.645 sexual assaults on victims aged over 13, 808 familial offences and 67 trafficking offences.
Now we need to move on the consensual offences where we have a breakdown by age as follows:
– sexual activity with a person under 13 years of age – 1773 recorded offences
– sexual activity with a person under 16 years of age – 4034 recorded offences
However what we don’t know here is anything about the ages of the other parties in these recorded offences. Some will be adults over the age of 18, where these offences assume a much greater severity in law, particular when it comes to the maximum sentence, which is 14 years imprisonment if penetrative sex is involved, other may well be under the age of 18 of even under the age of 16 – the maximum sentence there is 6 months imprisonment.
Because the law deems the consent of the victim to be irrelevant to the offence the scope of this offence covers everything from adults exploiting teenagers, which most would regard as a bad thing, to irate parents reporting their 15 year old daughter’s similarly aged boyfriend to the police because their pissed off that their daughter has become sexually active rather sooner than they’d hoped and there’s no way to separate out one type of offence from another, even though the latter is not only considerably less likely to result in a prosecution.
An unknown proportion of these offences are likely to relate to nothing more than fully consensual relationships between teenaged lovers of a similar age, relationship that no one in their right mind would deem to be either assaults or attacks, and never the tabloid’s inappropriate propensity for labelling any sexual offence where the victim is under 16 in terms of paedophilia.
What we can say is that the victim in all these reported offences was under 16, so notionally we have another 5.807 offences to add to our total of verifiably offence and a grand total of 17,873 recorded offence where we can be confident that the victim was a child so far as the law is concern, but even figure – which is over 5,000 less than the NSPCC’s headline figure comes with some pretty hefty caveats.
It also well worth noting that even if we make the unreasonable assumption that all of the 808 familial offences and 67 trafficking offence should be treated as offences against children this still leaves us with a balance of 4,349 offences to made up from the 17,645 sexual assaults on persons of any age above 13 just to get to the NSPCC’s figure. That’s almost a quarter of all sexual assault against over 13’s reported in England and Wales in 2010/11 that would have to be ascribed to victims ages 13 to 15 years of age just to make up the numbers.
One issue that was not reported in any of the major news outlets or mentioned in the press release by the NSPCC is problems associated with statistics on child abuse. All police recorded statistics on sex crimes (including sexual abuse of children) is affected by the problem of under-reporting.
This problem was acknowledged in the Home Office Report on Crime in England and Wales 2010/11:
“Police recorded statistics on sexual offences are likely to be more heavily influenced by under-reporting than the BCS [British Crime Survey] and therefore should be interpreted with caution.”
This document went on to say that analysis of the 2009/10 BCS self-completion modules showed that only 11 per cent of victims of serious sexual assault (not just cases involving children) told the police about the incident.
There are few reliable figures for under-reporting for any sexual offences however for rape, where we do have some reasonably reliable estimates, the figures show that just over four out of every ten rapes where the victim is female and under 16 are reported to the police compared to 3 out of every 20 rapes for adult women victims.
The likelihood then is that the figures mentioned in the papers today from the NSPCC’s findings underestimate the true number of victims of child abuse in this country. We contacted the NSPCC about this issue of under-reporting and they agreed that the real figure of the number of children sexually abused in Britain is probably higher.
Well yes, quite – but that doesn’t excuse the rather dodgy methods that the NSPCC has used to arrive at its headline figures.
This bring me at least to Full Fact’s conclusions which require an extra caveat or two, given what we’ve just seen above.
The figures on the level of child sex offences in Britain reported today are all accurate figures. In 2011 there were 23,097 reported child sex offences in Britain.
Well the figures are certainly accurate in terms of the way in which the NSPCC framed the questions it asked police forces via FOI but – and this is very big but – those questions do not accurate reflect the manner in which child sex offence are defined in law. As a result it would appear that data relating to the most offences included offence against adult aged 16 and 17 which has been innaccurately presented by the NSPCC and the press as offences relating to child.
Factor in the blanket use of terms such as ‘sexual assault’, ‘child sex attacks’ and – inevitably – paedophilia when referring to figures which include consensual offences and are also highly likely to include offences arising from consensual sexual relationship between teenagers and what you have are some very misleading headline claims feeding into the tabloid’s well established ‘Paedogeddon’ narrative.
Redemption for Full Fact, on this occasion, comes from the fact that they are, if nothing else, mindful to a fault of the caveats that need to be applied to all statistics that arrive via press release:
Most news outlets obtained their statistics from a press release by the NSPCC who themselves gained the information from 43 police forces across England and Wales. We have been able to verify most of this information, although we cannot account for possible discrepancies in how the FOI requests were interpreted by each force.
However, these figures should be taken with a pinch of salt because, like all statistics on child sex offences, they suffer from the problem of under-reporting. The likelihood is that the real figure of the number of child sexual offences is much higher than those mentioned in the news today, a statement which the NSPCC support.
And as we’ve now seen, that pinch of salt needs to be a rather larger one than even Full Fact suggest.
To be fair to Full Fact, its taken me a couple of hours of research and number crunching to put the data for this post together and sort out the analysis, and that’s from someone who’s already pretty familiar with SOA 2003 and the CJS and Home Office data that relates to it because of other posts I’ve written in the last couple of years, so one cannot really criticise Full Fact for not spotting the nuances in the NSPCC’s data and putting more time into checking it out.
Overall, they do a damn fine job of keeping tabs on the media while covering a hell of a lot of ground in a short space of time, so they’re bound to miss something every once in while, especially when it comes to something as complex as sifting the nuances of criminal justice statistics – at least it leaves something for the rest of us stats geeks to do.