Volvo lives up to image for crash safety: Government publishes analysis based on accident statistics for first time. Christian Wolmar reports

IF YOU WANT to survive a car crash, do not buy a Citroen 2CV or Dyane. If you are into big cars, avoid Vauxhall Carltons. Best of all, buy a Volvo.

For the first time, the Department of Transport has published its survey on which cars fare best in crashes. From an examination of 109,000 accidents involving two cars, the department's statisticians have compiled a list of those cars which offer drivers - not passengers, who have been excluded in the analysis - the best chance of avoiding injury or death in a crash.

The overall message is the rather obvious one that bigger cars tend to be safer for drivers than smaller ones. The worst performing car in the executive and luxury class, the Vauxhall Senator, has just about the same level of safety as the best car in the mini and supermini class, the Renault 5.

However, within categories there are substantial variations. In the executive and luxury class, the Jaguar XJ performs between 45 per cent and 60 per cent above the average for all cars, while the Senator is only between 5 per cent and 25 per cent above average.

In the upper medium range, Austin Ambassadors, Mercedes 190s and VW Passats stand out, while in the lower medium car category, it is newer Vauxhall Astras built between October 1991 and December 1992 which offer the best safety features.

Renault 5s and Peugeot 205s are the best in the small cars category while Citroen 2CVs and Dyanes, which are considered together since they have similar characteristics, are by far the worst car in all categories, averaging between 28 per cent and 41 per cent below overall average.

Volvo lived up to its much- advertised 'safe' image. While Volvos did not top any of the categories, three out of four of their ranges included in the survey had above average safety records, the best performance of any large manufacturer.

Robert Key, the roads minister, said in launching the report: 'Using this booklet could make the difference between life and death in an accident and motorists should take note of it before buying their next car.'

But he accepted the booklet dealt only with the prospects of avoiding injury or death after an accident had occurred, and not with the chances of avoiding an accident.

Department of Transport statisticians explained that it was impossible to consider the chances of being involved in a collision for each type of car, since different models had significantly different mileages. A spokesman said: 'You would have to assess the average exposure in terms of miles and mileage statistics are not collected.'

Manufacturers have criticised previous attempts by the department to assess the accident rates of different models because of the absence of mileage figures. The department admitted that the booklet gave only a partial picture of car safety by not including these accident rates.

The booklet does not consider the casualty rates for pedestrians and other drivers involved in accidents with a particular type of car. However, the booklet warns against certain features such as the increasingly popular 'bull bars' because they increase the risk of fatal injuries to pedestrians.

It says 'whilst at 20mph many children could survive an impact by a car, in theory a vehicle with a bull bar could kill at 10mph'.

The Royal Society for the Prevention of Accidents was critical of the exercise. A spokesman said: 'Our main concern is with safe driving, rather than safety features.'

Choosing Safety; HMSO; price pounds 2.50.

(Graphic omitted)