I READ your two recent letters on seat belts (March 7 and 21) with interest as I have just examined the evidence of their effects. You may be interested in the outcome. The prevalent, commonsense view is that seat belts reduce the number of fatalities and

I READ your two recent letters on seat belts (March 7 and 21) with interest as I have just examined the evidence of their effects. You may be interested in the outcome. The prevalent, commonsense view is that seat belts reduce the number of fatalities and serious injuries in road accidents. This is not what the figures show.

Seat belt legislation was introduced because research showed that, in any particular type of accident, people using seat belts were less seriously injured than those who did not. This is perfectly true.

However, using this evidence to introduce legislation assumed that the wearing of seat belts does not affect the behaviour of drivers. Hence, the number and type of accidents would not change.

This assumption is false.

In broad terms, drivers adjust their driving habits so that they feel the same degree of safety. This is known as risk compensation. The result is that they drive more boldly and have more accidents. If they have an accident, the injury to third parties is likely to be more severe because the vehicle is travelling faster.

The overall balance depends on subtle changes in numbers of accidents and injury statistics for both car occupants and other road users. Some safety devices appear to make matters worse (for example, anti-lock brakes), while others are effective, but usually less than was expected.

What appears to happen in the case of seat belts is that deaths and injuries to drivers reduce, but deaths and injuries to other road users increase by roughly the same amount.

In 1981, the Department of Transport commissioned a report that assessed this effect.

It was known as the Isles Report and concluded that "it [seat belt legislation] has not led to a detectable change in road deaths".

Similar conclusions have been reached in other countries. The Isles report was suppressed at the time but has subsequently been leaked.

Many people will dismiss this as nonsense. It is obvious, is it not, that seat belts save lives. If that is the way you think, I invite you to imagine driving with a sharp metal spike attached to the centre of your steering wheel. Might that not affect your driving, making you more cautious and alert?

Seat belts and many other safety devices have a similar effect but in the opposite sense. Because we feel safer, we drive slightly more boldly. Because we have more accidents, third parties are injured more often. Unfortunately, the evidence is quite technical and buried away in journals that are not freely available. However, I should acknowledge a book, simply called Risk by John Adams, which gives a very authoritative account.

I suspect your reader (March 7), who suggests that the police should have better priorities than enforcing seat belt law, has a point. I wear a seat belt because it is the law, but in the knowledge that by doing so I am slightly increasing your risk of being killed or injured.

MIKE GILDING, Home Farm Road, Houghton