Speed cameras 'have saved only HALF the lives ministers claim'

Speed cameras have saved far fewer lives than the Government claims, a study shows.

Instead of saving 100 lives a year, the true figure could be less than 50, they say.

'Speed cameras do reduce accidents but not quite as much as has sometimes been claimed,' said Dr Linda Mountain of Liverpool University's engineering department.

Official figures compiled by the National Safety Camera Programme compare accident rates before and after a speed camera goes up.

In the past decade, almost all cameras have been installed in notorious accident black spots. Police guidelines define these as stretches of road where there have been eight accidents in three years, including four serious crashes.

However, in a country the size of Britain, clusters of accidents are not always caused by dangerous roads or speed. Driver error, weather conditions or simple bad luck can also play a part.

If a camera goes up at one of these fluke sites, and accidents drop, the official statistics will claim that the camera has been successful.

But it is likely that the accident rate would have fallen anyway. The effect, known as 'regression to the mean', is well known to statisticians but not taken into account in camera statistics.

'If accidents fall dramatically after the cameras are in place, it could be that the cameras have reduced accidents, but some of the fall may simply show that a run of bad luck has come to an end,' Dr Mountain told the British Association science festival in Liverpool.

'The problem is that, although some parts of the road network are undoubtedly more dangerous than others, there is also a degree of randomness in where accidents occur.'

The last official figures, published in 2005, covered 4,100 cameras across Britain. They showed a 22 per cent fall in all accidents, and a 42 per cent drop in fatalities and serious injuries after cameras went up.

To find out the effect of regression to the mean, Dr Mountain looked at 215 speed cameras in more detail.

The official figures claimed a 25 per cent fall in accidents and a 50 per cent drop in serious injuries and deaths.

But her analysis showed the real figure was much lower.

'The big difference is that when you allow for the regression to mean effect, you find a 19 per cent reduction in accidents and serious accidents,' she said.

'I believe that cameras do reduce accidents, but not quite as much as is claimed.'

Officially, cameras are acknowledged as saving around 100 lives a year. But Dr Mountain believes the true figure is 'around 50'.