TCS Daily

Statistical Traffic Wreck

By Iain Murray - April 28, 2003 12:00 AM

Last week we heard on every news channel and read in every newspaper the disturbing news that more people had been killed in traffic accidents in 2002 than the previous year. The National Highway Traffic Safety Administration (NHTSA) had announced that 42,850 people had been killed in 38,356 fatal crashes, compared with 42,116 killed in 37,795 fatal accidents in 2001. There was much breast-beating and furrowing of brows at the news and the blame industry went into overdrive. The NHTSA itself provided much ammunition, and two of the blame industry's favorite targets were fingered: SUVs and alcohol. Yet a closer look at the details reveals that these figures are not the disaster they were portrayed as, and that there is one culprit that stands out above all else as the cause of most deaths.

To begin with, when we hear that more people have been killed than in a previous year, we must always remember that the American population is also climbing. More people in America means more people on the road. More people on the road means more opportunity for accidents, so it is not necessarily a surprise that more people have been killed in them. The question we need to ask is whether more people were killed per vehicle mile traveled in 2002 than in 2001. The NHTSA provides the answer: in 2002, there were 1.5 fatalities per 100 million vehicle miles traveled. That figure is identical to the 2001 figure, which in turn was the lowest figure in recent years (the number was over 2 until 1991, when it dropped below that level for the first occasion in modern times).

So our roads are actually as safe as they were last year, which is the safest they have ever been since the era of mass automobile ownership began. The consistent trend of our roads getting safer over the last decade is continuing. Moreover, when it comes to traffic injuries, the news is even better. Injuries per 100 million vehicle miles dropped from 109 in 2001 to 103 in 2002, a decrease of 5 percent. This is part of a consistent decline in injuries since 1995.

Nevertheless, the NHTSA provided figures that told reporters that SUVs, pickup trucks and vans were responsible for 59 percent of the increase in fatalities for vehicle occupants. SUVs, unsurprisingly, got the lion's share of the blame for the 500 extra fatalities in this category. Yet once again, this does not tell the full story. Just as the population of America has been increasing, the number of SUVs has also been increasing as they become more popular. The number of SUVs registered in the USA tripled from 976,000 in 1989 to over 3 million in 2001. There may be a safety issue with SUVs, but it is also possible that reckless drivers who would get into an accident anyway are now more likely than before to drive an SUV. It would therefore be useful to know the number of fatalities per SUV registered, but the NHTSA does not provide this information. It is also noticeable that the number of injuries sustained in SUV, pickup and van accidents declined from 861,000 to 848,000.

The NHTSA also stated that the number of alcohol related fatalities increased by 3 percent, from 17,448 to 17,990. However, when we look at the number per 100 million vehicle miles traveled, we find the figure increases by a tiny amount from 0.63 to 0.64. The rate has been increasing marginally from a low in 1999 following a decade long fall, but it is still well below the level of as recently as 1996. It is also interesting that the number of fatalities by those "impaired" by alcohol, in other words those who had taken a drink but were still legally allowed to drive (a blood alcohol content of less than 0.08), fell by 7 percent. The increase in alcohol-related fatalities came in those already recognized as intoxicated. There is certainly no case here for lowering the legal drinking limit further. Instead, these figures seem to confirm the well-established case that it is "hardcore" drunk drivers that are the real danger on the road and that policy changes should be targeted at them.

Alcohol played a role in 42 percent of traffic fatalities and SUVs, pickups and vans in about 35 percent. Yet the NHTSA findings pointed to a much more dangerous factor. Of all fatalities, 59 percent involved someone failing to wear a restraint like a seatbelt. Even a drunk driver in an SUV is much more likely to survive a rollover if he is wearing a seatbelt. Equally, a sober driver in a subcompact is more likely to survive if she is wearing a seatbelt when a drunken SUV driver crashes into her vehicle.

Not using a seatbelt is the single most dangerous thing a driver or other vehicle occupant can do. There is, however, no industry that makes money out of people not wearing seatbelts (except possibly for funeral directors). Instead, reporters have to blame their viewers and readers individually. It's much more convenient to accuse evil SUV manufacturers and the alcohol and restaurant industries. Yet because blame is focused elsewhere, fewer people will get the one most vital message that could help save their lives.

TCS Daily Archives