Why do measures to reduce speed get some people so annoyed? I ask this because of Mark McArthur Christie's intemperate article about speed cameras in these pages on 24 August. His views are simple, neat and wrong. Speed does kill. If we could cut speeds, we could reduce the avoidable cumulative death toll that we tolerate on our roads.

Let's get some definitions straight from the outset. Speed can be excess (breaking the posted speed limit) and inappropriate (driving too fast for the conditions). Here's an example of the latter. To drive a car at 70mph on the motorway may be legal. Driving at that speed in thick fog, however, is suicidal or murderous.

I accept that speed cameras catch only people committing excess speed offences. Yet these people are more likely to be involved in a crash. Ashton and Mackay (1979) found that there was a clear link between speed and accident severity: the change from predominantly survivable to fatal injuries among pedestrians happens between 30mph and 35mph. Hobbs and Mills (1984) pointed to the higher risk of injury to belted car occupants as speed rises. Those inside and outside the car are more likely to die if you travel faster.

Stradling concluded from a sample of Scottish drivers that those caught speeding either by camera or police officer were twice as likely to be involved in a crash. Those who break the speed limit regularly are "crash magnets" and need to change their behaviour.

Work from the Transport Research Laboratory has shown the link between speed, crashes and injuries. If average speeds drop by 1mph, crashes fall by 5 per cent. More specifically, the reduction will vary according to road type and average traffic speed from about 6 per cent for urban roads with low traffic speeds to about 3 per cent for higher speed urban roads and rural main roads.

It has also shown that accident frequency on urban roads rises with the increasing proportion of drivers exceeding the limit and with increases in mean excess speed. So if we can reduce the number of people breaking the speed limit and the speeds at which they travel, we can reduce deaths and injuries.

That's where cameras come in. The recent independent report into the first three years of camera implementation showed that speeds had fallen at camera sites by an average of 7 per cent, that the number of people breaking the speed limit at fixed camera sites had dropped by 71 per cent and that there were more than 100 fewer people killed at camera sites. I can't really see why Mark complains: if there is a law in place, surely we should comply with it.

He has a point about road deaths having risen last year. Here, however, he chooses to make 2+2=5 rather than look at the data. The two groups in which deaths rose were car occupants and motorcyclists. Among car drivers, we appear to have a growing incidence of non-seatbelt wearing among fatal crash victims. Some drivers have forgotten that belting up is more likely to save your life if you're in a crash. We need to repeat that message.

Motorcyclists are a tougher proposition. Twenty-eight per cent of motorcyclist deaths between 1997 and 2002 were in single vehicle accidents; 14 per cent of deaths occurred while overtaking, twice the level among car occupants. And motorcycling is highly seasonal.

There remains a big challenge to reduce the level of risk of death on the roads to that tolerated on the railways. That reduction will require political leadership and courage. To suggest that reducing speed plays no part defies research and reality.

Robert Gifford is executive director of the Parliamentary Advisory Council for Transport Safety

Search for used cars