I saw Malcom Gladwell’s take on SUVs the other day and was impressed. And I’m not the only one. But there is one glaring problem with it – the table that purports to show that SUVs aren’t safe has some problems. I’m reproducing it here:

Make/ModelTypeDriver DeathsOther Deaths Total
Toyota Avalon 
large402060
Chrysler Town & Country 
minivan31 36 67
Toyota Camry mid-size 41 29 70
Volkswagen Jetta subcompact47 2370
Ford Windstar minivan 37 3572
Nissan Maxima mid-size 53 26 79
Honda Accord mid-size 54 27 82
Chevrolet Venture minivan51 34 85
Buick Century mid-size 70 23 93
Subaru Legacy/Outbackcompact 74 24 98
Mazda 626 compact 70 29 99
Chevrolet Malibu mid-size 71 34 105
Chevrolet Suburban S.U.V.46 59 105
Jeep Grand Cherokee S.U.V.61 44 106
Honda Civic subcompact84 25 109
Toyota Corolla subcompact81 29110
Ford Expedition S.U.V.55 57112
GMC Jimmy S.U.V.76 39 114
Ford Taurus mid-size 78 39 117
Nissan Altima compact 72 49 121
Mercury Marquis large 80 43 123
Nissan Sentra subcompact95 34 129
Toyota 4Runner S.U.V.94 43 137
Chevrolet Tahoe S.U.V. 68 74 141
Dodge Stratus mid-size 103 40 143
Lincoln Town Car large 100 47 147
Ford Explorer S.U.V. 88 60 148
Pontiac Grand Am compact 118 39 157
Toyota Tacoma pickup 111 59 171
Chevrolet Cavalier subcompact146 41 186
Dodge Neon subcompact 161 39 199
Pontiac Sunfire subcompact 158 44 202
Ford F-Series pickup 110 128 238

Looks pretty authoritative, doesn’t it?

First off, the other deaths and therefore total death column is meaningless. What it measures is how often somebody besides the driver is riding. Automobiles are not made safe only for the driver, and what with the steering wheel like a blunt spear pointed right at the driver, you could argue that the driver sits in the most dangerous seat in any vehicle. So you should ignore that other death column, and concentrate only on driver deaths. It’s the only way to get an apples to apples comparison. Now the SUVs don’t look as bad.

Secondly, this table doesn’t take into account the driver. Young and old drivers are bad drivers. The Pontiac Sunfire may have such a poor record in part because it’s mainly driven by young hot rodders. The Lincoln Towncar may be less safe than the Ford Explorer in part because the drivers tend to be doddering oldsters who shouldn’t be on the road any more, not because the car is less crashworthy – and it might deliberately have lousy handling so as to give grandpa the feeling he hasn’t left his living room, which could affect its safety. And anyone only casually acquainted with America realizes that a different car models have different demographics – even with similar age ranges. It’s not only that a different age group drives mini-vans that drive subcompacts, but youths who want sporty (and thus drive more daringly) on a budget may prefer Sunfires to Sentras. 

Thirdly, the chart is per million cars, not million car-miles. So it doesn’t cover milage or how cars are driven (which sort of goes along with demographics). When I take my son to one of his activities, I’m often the only car in a sea of mini-vans and SUVs. Somehow, I don’t think there are all that many mini-vans on the road after the bars close on a Saturday night, the most dangerous time to be on the road.

Lastly, it lumps different models together, even though newer models may be much safer (or even less safe) than older ones.

Now I don’t think my objections mean SUVs are as safe as mini-vans (I own a mini-van and a subcompact), but I don’t think that the chart is conclusive, at best it’s suggestive.

I think you’re much better off looking at crash test results to get a handle on relative safety between vehicles.