BBC news story flagrantly manipulates statistics on school accident claims

A major news story on the BBC website this morning uses false comparisons and basic errors to create a highly misleading picture about the sums paid out for accident claims in schools. Far from revealing a ‘claims culture’, the figures actually show that payouts make up a tiny proportion of education budgets, and are not on the rise.

Screengrab BBC News home page 7 April 2017 with school payout story circled

The story gives the results of a Freedom of Information request on compensation payouts by local education authorities (LEAs). It claims that the figures “reveal an apparent doubling in school payouts from £1.65m in 2014 to £3.45m in 2016”. In fact this ‘increase’ is entirely explained by a single payout of over £2m after a pupil’s near-drowning led to severe, life-changing injuries.

But this statistical sleight-of-hand (which, in fairness, is mentioned in the piece) is not the most blatant manipulation of the figures. The story also states that “the total sum paid out… was £7.25m over three years, enough to fund the annual salaries of more than 320 newly qualified teachers for a year”.

Let’s walk through what this comparison actually does. It takes a total amount that LEAs have paid out over three years, and compares it to the salary cost of 320 of new teachers for a single year, clearly implying that the first could pay for the second. To which I am tempted to respond: why stop at 3 years? Why not multiply up to get a figure for 10 or 20 years, implying that payouts would pay for the salaries of thousands of teachers?

The only justification for this nonsensical piece of arithmetic is that it allows the piece to make an attention-grabbing comparison that grossly inflates the scale of the problem. It is terrible journalism, and I will be taking it up in a formal complaint to the BBC. (I have added this link in case any readers want to do the same.)

For anyone wanting a sense of the scale of the problem of accident claims, the BBC figures have another major drawback. They include payouts for abuse claims as well as accidents. The article states, for instance, that nearly half of the total paid by Kent County Council related to two historical sexual abuse claims from the 1970s.

I doubt the BBC wants to imply for a moment that payouts for child abuse are the result of ‘pupil mishaps’. So why include them in the statistics used in this piece?

Why this matters

Why does all this matter? It matters because – as I know from my talks and workshops up and down the country over many years – educators have real anxieties about being sued for accidents and injuries. And a key source of these anxieties is the relentless media narrative of a rampant claims culture.

Hard facts about claims help to give us a reality check. Especially when they are in a meaningful form, using relevant comparisons.

The irony is that, properly understood, the BBC figures are actually helpful in providing this reality check. I would argue that the real news story here is the opposite of that implied in the piece. Claims are, by any reasonable measure, low, and are not on the rise.

Here is my rough take on the BBC’s own figures. It leaves out the single £2million payout (as a roughly-once-in-a-decade outlier). And it just looks at accidents, not abuse.

Annual amount LEAs paid out on all compensation claims £1.5m Annual amount LEAs paid out on accident claims £1.2m Annual amount per LEA paid out on accident claims £24,000 Equivalent amount in annual teacher salaries 1 per LEA Equivalent annual amount per school £60

This comparison shows that each school paid out around £60 a year on accident claims, and each LEA around £24,000. Even if the figures were reduced to zero, the saving would equate to about a pound per week per school. Across a typical LEA, this would not even meet the full costs of a single additional teacher.

Let us not forget that some payouts are entirely justified. It is surely hard to argue with a claim from children who were injured when a heating duct fell on them (to take one of the BBC’s examples).

Even with the eyebrow-raising payouts (such as the cricket ball injury or the child falling off a table) it is worth mentioning that the devil may be in the detail. What sounds at first like a crazy decision to pay a claim can end up looking rather different. A good example is the infamous US Mcdonalds coffee scolding case. (Warning: this link contains details of injuries that may upset some readers.)

So here is my alternative BBC headline:

Mock BBC headline: Hard facts bust the myth of a 'compensation culture' of school accident payouts

My alternative headline for this story

Two morals

What morals can be taken from this story? First: there simply is not an epidemic of claims for school accidents. On any reasonable measure, the actual level of payouts is not a major problem. The fear of claims may well be a problem. But it is partly fuelled, I would suggest, by exactly the kind of media scaremongering shown in this BBC piece.

Second, the myth of the compensation culture is strong. So strong that supposedly respected news agencies are apparently willing to manipulate their coverage to fit it, even when it flies in the face of their own facts.

Notes on figures
* Based on a notional estimate that 25% of claims relate to abuse (the BBC figures provided do not allow for an accurate estimate)
** Based on 2012 government figures on numbers of LEA maintained schools (see tables 2a and 2b)

10 responses to “BBC news story flagrantly manipulates statistics on school accident claims

  1. This is great information to have especially since I’ve been told in recent weeks that things are so restrictive herein the States do to the fear of lawsuits regarding the safety of children. Also, great job filing a complaint as opposed to just discussing the matter!

  2. Can we bring the HSE into this, perhaps with another of their public ‘mythbuster’ pieces?

  3. Thanks Tim – I saw this article yesterday and was equally appalled. I’ve submitted a complaint as well. Your analysis has been really helpful here.

  4. Reblogged this on grumpysutcliffe and commented:
    Somehow doesn’t this just fit the steady erosion of our liberal culture? thanks Tim for blowing the whistle and alerting us.

  5. Thanks for this Tim. I have also submitted a complaint about this scaremongering

  6. Cracking spot Tim, good bloody work.

  7. Reblogged this on Bernard Spiegal and commented:
    Good, Tim. Needs nailing.

  8. The long term dangers of stressful, overprotected and sedentary childhood is more likely to damage children than the relatively few tragic accidents that have taken place!

  9. Thanks for the feedback – and also the shares via various social media. Several colleagues and contacts have stated that they have also submitted complaints to the BBC. I will report back on the outcome of my complaint.

  10. Thin Edge of the Wedge – The Problem of looking exclusively at liability payouts.
    Liability Pay Outs, particularly in playgrounds, are not the same or necessarily reflect the injuries or risk of injuries or their severity and should not be used as a surrogate for playground accident statistics. This is particularly true in a single payer or national health care system. Many involved in the provision of play consider these injuries are not as severe or frequent as children injured in automobiles or as pedestrians. That does not diminish that many of these injuries are preventable and should therefore be prevented.
    The European Playground Standards En1176 state in their introduction “Respecting the characteristics of children’s play and the way children benefit from playing on the playground with regard to development, children need to learn to cope with risk and this may lead to bumps and bruises and even occasionally a broken limb. The aim of this standard is first and foremost to prevent accidents with a disabling or fatal consequence, and secondly lesson serious consequences caused by the occasional mishap that inevitably will occur in the children’s pursuit of expanding their level of competence, be it socially, intellectually or physically.”
    The question really should be; do injuries with serious consequences occur? and do the performance requirements of standards actually prevent the risk of serious consequences or at worse have disabling or fatal consequences?
    To add clarity to what an injury with serious consequences is, ISO TC83 on Sports and other recreational facilities and equipment, published Technical Report ISO/TR 20183, Sports and other recreational facilities and equipment – Injury and safety definitions and thresholds – Guidelines for their inclusion in standards. This document defines a serious injury as “acute physical injury requiring medical or surgical treatment or under the supervision of a qualified doctor or nurse provided in a hospital or clinic and includes injuries such as burns, fractures, lacerations, internal injury, injury to organ, concussion, internal bleeding, etc.” Additionally a debilitating injury is “injury that diminishes or weakens the human body and a legacy of greater than one month and that could be categorized as abbreviate injury scale (AIS) of 3 (serious, but not life threatening).” The European General Product Safety Directive requires the prevention of serious risk to health, while the US and Canadian Consumer Product Safety Acts require prevention of serious injury.
    Collecting injury data is often difficult and can only truly be collected in conjunction with public health as the owner/operator of the facility or sports association where the injury took place fear litigation and therefore withhold that information. The issue is further complicated with the improvements in medical technology and the intake is no longer exclusively being the hospital emergency room(ER). For injury data collection, the US uses the National Electronic Injury Surveillance System (NEISS) to collect data at the ER, while Canada does this using the Canadian Hospitals Injury Reporting and Prevention Program (CHIRPP). In 2004 the US Consumer Product Safety Commission (CPSC) pointed this out in their report for 2002 to US Congress in that NEISS indicated well over 220,000 playground injuries treated in the ER, while medically treated were over 500,000 for the year. The US CDC also recently pointed out in a 2016 paper by Cheng, et al. that even at the ER intact the most obvious injury is statistically recorded and other less obvious and potentially more serious injuries such as concussions are only noted in the medical chart, but not recorded statistically causing this type of injury to be significantly under reported. Lastly there are jurisdictions that do not specifically record injuries in the playground, with the easiest explanation being that since there are no statistics, there must also be no injuries.
    Lack of statistics can bring a European researcher to a dead stop if they are not prepared to become creative and consider other options. The journey begins with a determination of the acceptable injury threshold. For arguments sake, take the stated goal of prevention of “injuries with serious consequences” from En1176 and the definition of serious injury from TR20183. The next consideration is to look around the world to find mechanisms and frequency of injuries on playgrounds. In lieu of national information, the statistics form the US and Canada, being 10 times smaller than the US, are instructive as they generally support each other from the point of view of number of injuries within the population and the major causes of injuries. Additionally children around the world tend to play and challenge themselves in the same way. Major injury mechanisms are the impact with the surface under the playground, 72-78%, while other injuries relate to the equipment and mostly fall into the failure of a product due to lack of maintenance. For falls it is possible to review the performance requirements for surfacing and the reality of protection from falls stated within the standards, while for maintenance a review of how seriously maintenance is taken within a standard is a good measure of the quality of maintenance that will be performed.
    As to the protection of children with regard to falls in playgrounds, the performance criteria are dismal and place children at risk. Military and Automotive data going back to the 1950s and to the cave dwellers when you consider a club over the head, show that impact to the head can result from serious injury to death. Although the detail is expansive and interesting to many, the key fact is that when a child lands on a surface within the “use zone” of the play equipment, they shall not experience an impact with a g value (force of gravity) greater than 200 or a Head Injury Criteria (HIC) value greater than 1000. This is tested using an instrumented headform, rather than a child, from a height stipulated by the standard and in Europe that maximum height for testing is 3 meters even though play structures in many playgrounds exceed 3 meters. Additionally for structures with barriers or guardrails, the drop test is performed at the deck height rather than the height of the barrier or guardrail the child must climb over to fall off the structure.
    So what does 200g or 1000 HIC really mean in relation to an injury outcome. Well the En1177 standard for impact attenuation says 1000 HIC is “a 3% risk of a critical injury (MAIS 5), a 18% risk of a severe (MAIS 4) head injury, a 55% probability of a serious (MAIS 3) head injury’.” Meanwhile CEN Technical Report CEN/TR 16148, Head and neck impact, burn and noise injury criteria – A Guide for CEN helmet standards committees indicates in table 4 that an impact value from 150 to 200 results in an AIS >3, while a value from 200 to 250 results in an AIS >4. For the same document table 5 indicates an impact value of 1000 HIC is a 10-15% probability of death, while a HIC value of 1500 is a 35% probability of death. US automotive data shows that a value of 200g is a 10% risk of skull fracture and more recent data looking at the incidence of concussions for National Football League players 80g is a 50% risk of concussion and concussion occur at 98g ± 28g. Obviously, for Europe, the performance requirement of the standard are well outside the injury threshold stated in the introduction or the legal requirements.
    The next area of investigation is the issue of maintenance as it relates to just keeping the playground within the performance requirements of the EN1176. This is important because playgrounds are dynamic outdoor environments subject to sun, rain, wind and worse children. The need for playground maintenance is critical and for the Europeans there is the En1176 part 7 that basically recommends maintenance through the multiple use of the word “should” rather than “shall”. At least the Canadians require daily/weekly, monthly and annual comprehensive inspections along with the prior to use inspection before children are allowed to play in a new playspace. Failure to perform inspections, surface testing and maintenance places children at risk for failure of equipment and impact with surfaces that are performing worse than allowed in the standards.
    To come full circle then it is not the liability claims and settlements, as they point to massive failures, but the injuries that are treated and released that are of concern from the point of view “is the health care system being asked to pay for the mistakes of others?”. In a “single payer”, “national healthcare” or “mandatory insurance” system a child with an injury sustained in the playground or anywhere else is treated and released and where necessary rehabilitation provided at no cost to the child or caregiver/parent.
    The question therefore is “Are playgrounds putting undo financial strain on the healthcare system for preventable injuries?” Now there’s a research topic worth looking into. The Cheng et al. paper would suggest that with regard to head injuries the answer is yes. Playground equipment recalls in the United States that have been voluntarily agreed to with the CPSC would also suggest yes.

Leave a comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s