M4 camera effects in doubt
Misleading stats and downright lies
M4 camera protest, May 2005
The SSCP claims that cameras have cut crashes on the M4 by between 44 per cent and 57 per cent in the year since they were installed. The Association of British Drivers (ABD) said it believes this crash reduction is "a remarkable achievement" given that fewer than 15 per cent of crashes on this stretch of motorway had excess speed as one of their causes, many of these not involving exceeding the speed limit.
"There's something very odd going on here," said the ABD's policy director Mark McArthur-Christie. "We have a speed camera partnership claiming credit for reducing accidents that weren't even caused by speeding. This seems disingenuous at best."
Selective statistics
The ABD said it believes the partnership is being extremely selective in the statistics it chooses to report.
McArthur-Christie said: "The camera partnership is taking credit for falls in deaths -- but the fatality rate from 2005 was identical to that in 1999 and 2000 -- despite the camera partnership only starting motorway operations in 2005." Serious injuries are down - but they are also down across the road network as technological developments like airbags and seatbelt tensioners save lives.
The ABD analysed the crash statistics and concluded that inattention, failure to judge other drivers' speed and changing lanes without looking were the biggest killers on this stretch of road (accounting for 75 per cent of fatal crashes), not speed. "This makes it all the more remarkable that the partnership is now claiming credit for crash rate reductions", said the ABD.
McArthur-Christie said: "We need to be careful not to attribute success to the wrong thing. The presence of cameras may be forcing drivers to pay more attention on this stretch of motorway, but the threat of prosecution and speeding tickets will not improve safety. If there's a genuine reduction in crashes it's because drivers have woken up and are more attentive - not because they're sticking to the limit."
Safe Speed's view
The Safe Speed road safety campaign weighed in with detailed analysis of the claims. According to Safe Speed, the figures do not support the claims:
- There's no good reason to expect speed cameras to reduce M4 crashes. Figures obtained by the Association Of British Drivers for M4 crashes revealed that 15 per cent had 'excessive speed' as a contributory factor. Many 'excessive speed' crashes do not involve exceeding a speed limit. So even a 'perfect' change in driver behaviour caused by the cameras should not be expected to reduce crashes by more than 15 per cent.
- Selection bias / regression to mean effect. It is a fact that we tend to apply road safety treatments to places where there have recently been high casualty numbers. This leads to a selection bias effect because some of those high casualty numbers will be due to nothing more than random chance. With no treatment applied, abnormally high casualties tend to reduce anyway. This effect is also known as 'regression to the mean' (RTTM), and is very well understood in road safety circles. Recent Department for Transport figures suggest that RTTM accounts for three quarters of the benefit that has previously been claimed for speed cameras.
- Too soon to tell. Road crash figures are subject to considerable random variation. It is unprofessional and illegitimate to make claims based on a single year's figures. Examples of random variation appear strongly in the figures released by Wiltshire Camera Partnership. e.g. between 2003 and 2004 personal injury collision reduced by 37 per cent from 261 to 165. There were no cameras on the M4 at the time.
- Cherry picking of comparison groups. By choosing favourable years or casualty groups the changes can be exaggerated to show a desired effect. Other figures -- for example M4 casualty figures for Wiltshire in 2000 and 2001 -- are unpublished. Even within the published figures there is ample room to spin the story. For example we could say: between 2001 and 2002 fatal and serious casualties increased from 18 to 31 - an increase of 72 per cent. This is just as valid as some other comparisons that are being made.
- System trends. During the period in question we have tended to see national reductions in injury and serious injury collisions. So if the M4 in Wiltshire was typical of UK roads we would have expected to see the numbers come down anyway, irrespective of any local treatment. This has not been allowed for and clearly exaggerates the benefit being assigned to the cameras.
- Local conditions. Changes in local conditions have a marked effect on accident rates. For example, large long term motorway road works would be expected to lead to a reduction in casualties.
- Alternative routing. Traffic chooses a different route because of the cameras and have their crash elsewhere. And in the case of the M4, the common alternative is M3/A303. Some of the apparent reduction may be due to traffic choosing a different route to stay away from the M4 cameras.
Exaggerated benefits
According to Safe Speed, these issues tend to exaggerate the benefits claimed for M4 speed cameras. "Until these issues are accounted for we will not know if there has been a genuine accident reduction on the M4 or not", said campaign founder Paul Smith.
Smith said: "Analysis and interpretation of road accident statistics requires considerable care and scrupulous honesty because road accidents are rare and exceptional events. It is necessary to separate random effects from real changes. I am absolutely incensed that, once again, a camera partnership is making outrageous and unjustified claims for their cameras based on insufficient data and insufficient analysis. It is in fact an object lesson in 'how to lie with statistics'.
"It may not be immediately obvious, but misleading road safety statistics are actually deadly. They cause life saving resources to be allocated to the wrong safety factor, and we therefore miss the opportunity to save lives.
"These figures do not show that speed cameras have made the M4 a safer place. Instead they show that the Wiltshire camera partnership is unscrupulous when it come to promoting themselves."
'We've had no effect on crashes'
Meanwhile, partnership manager Dave Frampton, speaking on BBC Radio Swindon this morning, admitted that there was no evidence whatsoever to suggest that their camera vans had caused a reduction in crashes on the M4.
Smith said: "This is typical of the cowboy behaviour of camera partnerships. They claim success at every opportunity based on no sound evidence whatsoever. Every lie or exaggeration distorts road safety priorities and risks increasing danger on our roads.
"After nearly 14 years of speed cameras on UK roads, there is precious little evidence of any benefit, despite £800,000,000 in fines issued to around 13 million motorists. Road deaths are NOT falling as expected and we're now the slowest improving country in Europe."
Links
Forums | Speed, Plod & the Law | Top of Page | What's New | My Stuff