Quantcast

"I drive a SUV because I feel safer in it"

Toshi

Harbinger of Doom
Oct 23, 2001
38,031
7,550
[an original post by Toshi]

:weee:

Introduction

I've heard the above line from the title of this post from quite a few people, usually smaller women, who drive SUVs. The sense of height and mass feels safer, apparently. Are these feelings supported by the data? Looking at the Insurance Institute for Highway Safety stats on Driver Death Rates by Vehicle Make and Model, most recently updated in 2007, can shed some light.

First, some shortcomings: The data were last published in 2007 so many recent vehicles, such as our 2nd generation Prius, are not included amongst them. Confounding driver characteristics--Corvettes and DeVilles are driven by different types of people!--are also in the mix, although the IIHS has tried to control for them. Finally, statistics about vehicles driven by people in the aggregate may not be applicable to you and your individual situation. That said, these are the best data that I know of.

Let's start with the intuitively obvious: smaller cars and trucks generally have higher driver death rates than larger vehicles. Here's the table that illustrates this:



MV: multiple vehicle collision; SV: single vehicle collision; roll: rollover

End of story? Hardly. Treating a Mercedes-Benz, a Toyota, and a Kia of the same curb weight identically is doing a disservice to all. Drilling down to the level of individual models is warranted. Below are some examples of individual models' statistics that prove or disprove conventional/folk wisdom:

1) "My pickup truck is safe because it's so large. Don't get in my way or I'll crush you!" Verdict: FALSE.

Pickups are less safe as a group than the average vehicle. (The average vehicle has has 79 fatalities per 1,000,000 registered vehicle-years, from hereon referred to simply as "deaths".) Some of the most popular pickups are even more dangerous than the group's average. In particular, the old-style pre-2004 Ford F-150 is/was particularly egregious, with 118 deaths (confidence interval: 95-141). Telling, too, is the makeup of these 118 deaths, with 89 of them being single vehicle collisions presumably related to poor handling secondary to the vehicle's huge size, and 60 of them being rollovers, for which no further explanation is necessary.

What about the F-250 and F-350, then? They're undeniably bigger than the F-150. Are they safer? No. They "boast" death rates of 122 (CI: 101-143) and 103 (CI: 79-128), respectively, with the vast majority of deaths being in single vehicle collisions and from rollovers. Do less pickup drivers die in multi-vehicle accidents? Yes--conservation of momentum at work. Enough extra drivers die in single vehicle collisions and rollovers, on the other hand, to make the vehicles substantially less safe than the average.

Moral of the story: Don't drive a F series pickup if you're after safety. The offset 40 mph crash tests of a 2002 Mini Cooper and a 2001 Ford F-150 shown below further illustrate that size does not necessarily equate to safety, especially when the mass of your vehicle meets an immoveable object:




What I can't explain is the disparity of the data for the sister SUV models to the F-150 and F-250, the Expedition and Excursion. The Excursion follows its F-250 brethren closely, with a poor 115 death rate, again largely comprised of single vehicle collisions and rollovers, but the Expedition actually does quite well with 31 deaths. Go figure.

2) "You get what you pay for." Verdict: TRUE.

The conventional wisdom is that if you buy a "cheap" car, say, a Kia, then you'll get something that's less safe than an equivalent "good" car such as a Lexus or Mercedes-Benz. The data do support this assertion.

First off, the case of the Kia: The Kia Rio has absolutely awful safety statistics. It has 175 deaths, mostly multi- and single-vehicle collisions but also with a fair number (35) of rollover fatalities. (As a side note, these statistics, which have carried over several model years, are why I elected to not buy a Rio myself several years ago despite the favorable price and good warranty.) Its standard Japanese competitors in the small-car arena, the Honda Civic and Toyota Corolla, have much lower albeit not spectacular death rates of 74 and 79. I'll still take a factor of 2 reduction any day of the week, thanks.

On the flip side of the equation, some brands, such as Mercedes-Benz, seemingly are deserving of their safety-conscious reputation. Although Volvo has made much of its safety through its advertising campaigns it's actually Mercedes that been at the forefront for many years, with the first series production airbags, for instance. Although demographic factors may skew the data, they presumably skew the data more or less uniformly throughout a given segment, and in the "luxury car" realm Mercedes is up at the top along with BMW and Lexus:



Note the excellent figures for the midsize Lexus ES (18), midsize Mercedes C class (24), large Mercedes E class (14), large Lexus GS and LS (29 and 30), very large BMW 7 series (11), and very large Mercedes S class (21). Again, the global average across all vehicle sizes and types is 79, so being below 30, let alone 20, is phenomenal. Lest one assume that this is merely an artifact of vehicle size, consider the old-technology Cadillac CTS, Buick Park Avenue, and Lincoln Town Car with their 81, 85, and 91 deaths.

It seems as if you do get what you pay for. If I was a luxury car buyer I'd be barking up the imports' tree without a question, and I don't include the Koreans amongst the Japanese and Europeans as of yet.

3) "It just seems like I'm safer in my SUV." Verdict: Neither true nor false, as it varies widely by model.

Finally, we address the now-old saw: "I just feel safer in my SUV." Well, are you safer in reality? It depends widely on the model.

As one example, my mother is neither safer nor less safe now that she's out of her SUV. Her Expedition actually is one of the better ones out there, with 31 deaths and only 11 rollovers. Her current Prius isn't listed in the IIHS data, but through extrapolation from NHTSA data could be expected to have a death rate between 35 and 45, about half the fleet average. Given that her newer Prius has many more airbags, stability control, and better HID lighting than the older Expedition, the on-paper difference might be even slimmer. I'm sure she doesn't mind going from 15 mpg highway mileage (observed) to 50 mpg, in any case!

Another example that's basically a wash is of a hypothetical soccer mom going from a (fwd) Honda Civic to its sister vehicle, an AWD Honda CR-V, for better snowy day abilities. The Civic has 74 driver deaths, mostly in multi-vehicle accidents where its low mass puts it at a disadvantage, whereas the AWD CR-V has 67 driver deaths, more evenly distributed between single-/multi-vehicle collisions and rollovers.

A less gas conscious soccer mom might move from a Ford Taurus to a Chevy Suburban. Again it'd be a mixed bag, as the 78 fatalities of the Taurus drivers include only 14 rollovers while the 66 fatalities of the Suburban drivers include 44 single vehicle collisions and 33 rollovers. Gain in one area, lose in another, and lose a bunch of fuel economy while at it…

Cherry-picking an unsafe SUV yields dramatic results: Going from pretty much any vehicle on the market save for a 350Z, RSX, or Kiato a 2-door Chevy Blazer would be a downgrade, as the Blazer's 232 deaths--134 from rollovers alone!--is the worst record out there. Other models to avoid include a few SUVs but a lot of "cheap" small cars:



Picking more carefully and avoiding lemons like the Blazer, however, one might actually see a very significant gain in vehicle safety going from a car to a SUV, as some of the safest models around in absolute terms are mid-size and large SUVs. Going with any of these mid-sized 4WD/AWD SUVs is probably a good choice, safety-wise, and if the Highlander Hybrid model is picked, might even not be that bad at the gas pump!



Conclusions

Conventional wisdom dictates that small cars are less safe than large. As a general rule this actually does hold true. However, there is so much individual variation between makes and models to make this rule nearly useless. Drive a 2001 F-150? You're actually much less safe than that Civic you peer down at disparagingly from aloft. Drive a Highlander? You're actually doing better even than the rich schmuck in an S Class in the next lane over.

Use the data. Choose wisely. Avoid the vehicles on the most-unsafe list. Most importantly, drive safely. It's difficult to quantify the effects of a safe driver, but it's likely on the same order of magnitude as the differences between most of these vehicles.
 

Silver

find me a tampon
Jul 20, 2002
10,840
1
Orange County, CA
Use the data. Choose wisely. Avoid the vehicles on the most-unsafe list. Most importantly, drive safely. It's difficult to quantify the effects of a safe driver, but it's likely on the same order of magnitude as the differences between most of these vehicles.
I can't find a source for it, but I remember they had some weird accident rates between a Dodge/Plymouth badged cars some years back. They couldn't figure it out, until they looked at the demographics of the buyers-and it turned out that the Dodge was being bought by a significantly younger group than the Plymouth. The Plymouth was much safer statistically, even though they were the same car.

Could all be BS though, since I can't find it....:rofl:
 

Toshi

Harbinger of Doom
Oct 23, 2001
38,031
7,550
I can't find a source for it, but I remember they had some weird accident rates between a Dodge/Plymouth badged cars some years back. They couldn't figure it out, until they looked at the demographics of the buyers-and it turned out that the Dodge was being bought by a significantly younger group than the Plymouth. The Plymouth was much safer statistically, even though they were the same car.

Could all be BS though, since I can't find it....:rofl:
There are a few similar pairs of cars in the 2007 IIHS data:

Mercury Sable 71, Ford Taurus 78. Ok, not impressive.
Ford Crown Vic 45, Mercury Grand Marquis 75. Here we go…

One could even argue that the Infiniti G35 sedan (11 deaths) and the Nissan 350Z (193 deaths!) are at least cousins under the skin… Similarly we have the Toyota Camry (55) and Lexus ES330 (18!).
 

dante

Unabomber
Feb 13, 2004
8,807
9
looking for classic NE singletrack
What I find interesting is that for cars, the # of deaths per multiple vehicle accident are all higher than SV accidents... And in a MV accident, there's a 50/50 chance that the accident wasn't your fault (ie, someone crashing into *you*).

For SUVs, the numbers are reversed (except for the smallest size). Sure, there's still a high number of MV deaths, but the number of SV deaths (ie, driver acting like an idiot and driving off the road) is higher.

Can we ban SUVs yet?
 

mattmatt86

Turbo Monkey
Feb 9, 2005
5,347
10
Bleedmore, Murderland
I feel like the stupidity of the driver could come into play. This list of cars is pretty much a list of what I always see the worst drivers in.
Agreed. This has me wondering about the 350z - g35 comparison Toshi brought up.

Me thinks that despite being essentially the same vehicle, there are more deaths in 350z's because they are cheaper than g35s.

Cheaper = younger more Fast and Furious type drivers = more reckless driving = more deaths. Or am I drunk?
 
Last edited:

kazlx

Patches O'Houlihan
Aug 7, 2006
6,985
1,957
Tustin, CA
I think there are so many other factors that those statistics really don't mean crap. Types of buyers, purchase prices of cars, etc.
 

Icantdrive65

Monkey
Mar 21, 2005
609
1
Chinquapin fire road
When my 98 Tacoma got totaled, I was doing probably 40-50, and the cab stayed intact. I was shaken, not stirred, bent, not broken in the accident. This is precisely why I bought another Tacoma.
Well, I better not tell you why my friend who crashed his 2001 Tacoma didn't buy another one.

The nature of the accident makes all the difference in the world. Lots of newer cars are engineered to pass specific safety tests, but may be more vulnerable in other types of accidents.
 

eaterofdog

ass grabber
Sep 8, 2006
8,189
1,431
Central Florida
SUVs and trucks tend to roll. Due to my commute, I see accidents on the interstate a lot. When cars fvck up at 80, they spin into the rail, tow truck comes, everybody goes home. When a truck/SUV breaks loose, they roll in a very vigorous manner that absolutely destroys the entire vehicle. And if anybody is not belted in, they get launched like they came out of a cannon.
 

Toshi

Harbinger of Doom
Oct 23, 2001
38,031
7,550
I see no Subaru's[sic] in there.
Full dataset is linked from the first paragraph of the introduction: http://www.iihs.org/externaldata/srdata/docs/sr4204.pdf

Subaru Legacy sedan: 40 deaths, pretty much evenly distributed between the categories
Subaru Legacy wagon: 50 deaths, basically the same as the sedan but with 10 more single vehicle deaths
Subaru Forester: 53 deaths, basically the same as the wagon

Yes, this would support that these Subarus are safe/are driven in a safe manner, albeit not to the extremes of the BMWs, Mercedes-Benzes, and Lexuses of the world.
 

Toshi

Harbinger of Doom
Oct 23, 2001
38,031
7,550
What's your numbers about motorcycles?
That's an interesting question. Here's part of the puzzle: ABS helps. A lot.

http://www.iihs.org/research/topics/pdf/r1110.pdf

Comparing like vs. like--same model with ABS and same model without ABS, presumably bought by otherwise-similar people--they found:

motorcycles [in their study] with ABS had a fatal crash involvement rate 37 percent lower than that for their non-ABS versions during the study years. This difference was statistically significant from [the null hypothesis of no difference] at the customary [p <] 0.05 level. Thus, there is considerable confidence that ABS is preventing fatal crashes among motorcyclists. This confidence is bolstered by the fact that a separate analysis of insurance collision coverage losses among crashes of all severities also shows a statistically significant reduction in crashes of about 22 percent for motorcycles equipped with ABS (HLDI, 2009). These results provide confirmatory evidence of the expected benefit of ABS from engineering principles, test-track trials, and a crash reconstruction analysis.
Here's another piece of the puzzle: driver age. http://www.iihs.org/research/topics/pdf/hldi_motorcycle_antilocks.pdf . They write:

Driver age was a significant predictor of motorcycle collision overall losses. Estimated overall losses for rated drivers 24 and younger were 113 percent higher (p<0.0001) than those for rated drivers ages 25-39 (reference category). Rated drivers ages 40-64 and 65 and older had slightly lower estimated overall losses, but these estimates did not reach statistical significance.
Finally, in broad strokes, the faster the bike, the higher the losses (and presumably rider death): http://www.iihs.org/research/qanda/motorcycles.html

Supersport motorcycles had the highest overall collision coverage losses among 2005-09 model bikes, almost 3 times higher than losses for touring motorcycles and almost 5 times as high as cruisers, according to data from HLDI. All 10 motorcycle models with the highest losses were supersports.
 

bizutch

Delicate CUSTOM flower
Dec 11, 2001
15,928
24
Over your shoulder whispering
We did an awesome sample test of ramming a 70's solid steel Monte Carlo into a brand new Monte Carlo drivers side front corner to corner head on. The old one impaled its test dummy and would have done the same had there been dummies in the back too. The cabin of the new one was completely intact and the dummy undamaged.

Oh, and the cost to insure a WRX...one of the highest of any car on the road.
 

stoney

Part of the unwashed, middle-American horde
Jul 26, 2006
21,519
7,065
Colorado
Oh, and the cost to insure a WRX...one of the highest of any car on the road.
Teenagers + Fast Car = Tree

The insurance on my Forester XT is almost half of a WRX, and it's 95% of the same car. And it's faster.
 

Jeremy R

<b>x</b>
Nov 15, 2001
9,698
1,053
behind you with a snap pop
Most of that is just common sense.
Take the Kia Rio for instance.
I would take a guess that the car itself is not that much more dangerous than comparable models.
But the people who buy a Kia Rio are poor decision makers in life.
The same person who says, "Hey, a Kia Rio is the car for me." is gonna be the same person who says, "Hey that 18 wheeler is bearing down on me at 75mph, but I think I can pull out in front of him if I step on it because of my new custom muffler. Its loud."
 

Toshi

Harbinger of Doom
Oct 23, 2001
38,031
7,550
Most of that is just common sense.
Take the Kia Rio for instance.
I would take a guess that the car itself is not that much more dangerous than comparable models.
But the people who buy a Kia Rio are poor decision makers in life.
The same person who says, "Hey, a Kia Rio is the car for me." is gonna be the same person who says, "Hey that 18 wheeler is bearing down on me at 75mph, but I think I can pull out in front of him if I step on it because of my new custom muffler. Its loud."
Why are its drivers dying at twice the rate of Corolla and Civic drivers then? Presumably the same frugal people are cross-shopping those vehicles. I was, before we bought our CPO Fit...