Skip to content

Lies, Damned Lies, and Statistics

July 29, 2009
Accurate Statistics Are No Accident!

Accurate Statistics Are No Accident!

Manchester, England – Not a good week to be a statistician working for the UK’s Department of Transport (DfT) as yet another statistical chicken comes home to roost.

A report by the UK Statistics Authority (UKSA) has accused the DfT of underestimating the number of people seriously injured in traffic accidents. The UKSA is so concerned by this issue that it has warned the DfT that its accident casualty figures will no longer be recognised as official unless their reliability is improved by November.

Now it’s difficult not to feel a wee bit smug on this one, although I’d be the first to admit that road traffic accidents are not an area where smugness should really be applied. However when you consider the fact that myself and others have on numerous occasions pointed out to the DfT that there is an issue with the voracity of the data used to compile the accident stats I think I’m allowed just a nanosecond of smugly.

This does matter particularly as it appears from UKSA’s report that twice as many people are being injured or killed as the DfT figures are reporting. The figures on road accidents are very sensitive because the UK Government insists that, thanks in part to the speed camera programme, it has already met its target of reducing the number of people killed and seriously injured to 60 of the per cent of the annual average in the mid-1990s. Not to mention the fact that the amount of resources, i.e. cash, that finds its way down the food chain to our hard pressed front line Road Safety Officers is actually based on the need created by the DfT accident stats.

All this confusion over the accident stats is caused by the fact that the UK has two systems of data collection. STATS 19 are a form filled in by police officers either at the scene of an accident or when an accident is reported to the police by other means. In the UK there is no need to report accidents other than injury accidents to the police thus STATS 19 has a tendency not to pick up minor shunts even when there may be an injury. On the other hand HES (Hospital Episode Statistics) are directly related to hospital admissions and are collected at Accident and Emergency Centres, they thus record nearly all road accidents even where very minor injuries are concerned.

Is this of interest to bikers? In some ways yes. I’ve always been fascinated by the low level of bicycle accidents in the UK, considering the number of people who cycle the accident rates are very low. This is obviously grist to the mill of the cycle lobby and give much pleasure to the DfT funded QANGO Cycle England, who frequently throw the ‘look how safe cycling is’ stats card  in the face of the UK motorcycle lobby. I don’t know why they do this, insecurity perhaps.

Now I’ve actually compared the English STATS 19 and HES data for both motorcycle and bicycle accidents and have discovered that there is a huge under reporting in relation to bicycle accidents. The motorcycle accident data is a reasonable match, quite understandable as most motorcycle injury accidents get reported to the police, where as cycle accidents don’t. People just fall off bicycles and go to hospital. Based on the HES data it’s true to say that in some types of injury more cyclists are being hurt than motorcyclists. Ok usage is a factor but it’s a fact that the overall accident figures for motorcycles and cycles are much much closer together based on HES than on STATS 19.

No surprise then that DfT and the cycle organisations like to use the STATS 19 data and do their best to discredit the HES data. Now please don’t get me wrong here, I’d like us to reach a situation where we have zero accidents. But if we are going to have a debate on accident stats and use said stats to support the promotion of one mode, cycling, over another, motorcycling, let’s use the most accurate and fact based data available and not data sets that very handily support a Government transport policy.

© Back Roads Rider 2009

One Comment leave one →
  1. Simonwb permalink
    August 2, 2009 7:32 pm

    Could I suggest a little data exercise that might throw up an interesting picture?

    Using the established ratio between Stats19 and HES that you have for motorcycling, could you estimate the true(r) numbers of cyclists accidents by applying this ratio to the cyclists HES? And then maybe factor in MKVT for cyclists? Shake vigorously and compare to motorcycling. Just a thought…

Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s

%d bloggers like this: