Speed cameras: do they cause fewer accidents and road deaths?

Thames Valley Police today announced their decision to switch back on their array of speed cameras in Oxfordshire, citing as their reason "an increase in fatalities and the number of people seriously injured" since they were switched off last year.

However the relationship between speed cameras and road safety has been a contentious one in the past, and is a topic that Full Fact as investigated time and time again.

When Oxfordshire took the decision to turn off their speed cameras in July last year, we looked at claims that the decline in accidents on Britain's roads had actually slowed since cameras were first introduced in 1992.

However it is difficult to draw this conclusion from the statistics alone. As a spokesperson for the Department for Transport told Full Fact:

"When you consider the range of factors that have contributed to the long-term improvement in road safety, from social changes to drink driving legislation, it is difficult to draw any useful conclusions with any confidence on speed cameras."

Furthermore he pointed out that it "would be much more useful to look at the number of accidents recorded at camera sites."  This is something today's release from Thames Valley Police addresses, finding that in the six months after they were switched off, 83 people were injured in 62 accidents at the site of fixed cameras. This compares to a figure for the same period the year before of 68 injuries in 60 accidents.

This problem of tracing causality is one we also encountered when tackling the Daily Mail's suggestion that there had been a 14 per cent fall in road deaths for a three month period in which cameras were not used.

After Swindon turned off its speed cameras in 2009, the local borough council produced statistics which many media outlets suggested showed that there had not been a subsequent rise in accidents.

However concerns were raised by Full Fact that the six month study might be too short to draw conclusions from. When we put this to the council, they agreed, stating: "We've always said statistically, it's going to take a long time - possibly up to two years - before you can drawn any meaningful conclusions from these statistics relating those particular site because it's such a small sample."

These potential pitfalls are all things readers may wish to bear in mind when considering this latest speed camera story.