What kind of car is the safest to drive?

December 27, 2005 11:51 AM

What kind of car is the safest to drive? There are lots of studies on which cars are best in a crash, but I suspect that some cars crash less often than others, so I am interested in statistics that take the rate of accidents as well as crashworthiness into account.

My girlfriend is shopping for a new (used) car and her sister, who often has a skewed view of risk, is insisting she get an SUV because it is bound to be safest in an accident. I suspect that while an SUV might be better in a crash, they might crash more often due to longer braking distance, higher rollover risk, and generally being less manuverable. Can anyone help with death or injury statistics by model or make of car? Are there any other indicators that would be useful to examine, such as insurance rates?

posted by procrastination to travel & transportation (24 comments total)

Have you looked at Consumer's Reports statistics?
posted by k8t at 11:58 AM on December 27, 2005

I suspect that some cars crash less often than others
This could be due to the type of driver in the car.
Car A has more accidents than car B not because its more dangerous but because 18-25 year olds prefer it, while the 40+ accounting crowd likes car B.
posted by MrMulan at 11:59 AM on December 27, 2005

I'd recommend this article by the New Yorker's Malcolm Gladwell. He discusses how the SUV's reputation for safety is a prime selling point but not, in fact, correct.
posted by docgonzo at 12:01 PM on December 27, 2005

you're better off getting her defensive driving lessons at a racing school than you are arming her with a 6000-lb juggernaut and a Frappuchino.
posted by kcm at 12:01 PM on December 27, 2005

The government puts out a crash report, I remember the VW Jetta being one of the top three in the latest. But, as kcm already mentioned, investing in a defensive driving course is going to be money well spent, no matter what kind of car she ends up driving.
posted by odinsdream at 12:08 PM on December 27, 2005

I agree with defensive driving school. You can go around in your armored-clad vehicle trying to feel secure, but it's the person behind the wheels and how they react to emergency situations that's the key. No amount of safety equipment is going to help you if you can't control a blow-out at freeway speeds; or get into a oversteer situation and don't have the proper instinct on how to react.

If you still need to comparison shop based on safety, head over to NHTSA's site.
posted by jaimev at 12:27 PM on December 27, 2005

The Malcolm Gladwell article (from docgonzo's comment) also discusses Consumer Reports "accident avoidance" testing, and has a table of fatality rates for specific models. Excerpt:

Drivers of the tiny Jetta die at a rate of just forty-seven per million, which is in the same range as drivers of the five-thousand-pound Chevrolet Suburban and almost half that of popular S.U.V. models like the Ford Explorer or the GMC Jimmy. In a head-on crash, an Explorer or a Suburban would crush a Jetta or a Camry. But, clearly, the drivers of Camrys and Jettas are finding a way to avoid head-on crashes with Explorers and Suburbans. The benefits of being nimble—of being in an automobile that's capable of staying out of trouble—are in many cases greater than the benefits of being big.

posted by mbrubeck at 12:34 PM on December 27, 2005

The article docgonzo posted discussed some research on exactly what I was interested in, and I was able to find the original academic papers on the topic. This paper had exactly what I was looking for.

Thanks, AskMeFi!
posted by procrastination at 12:43 PM on December 27, 2005

Remember advice from Mr. Myagi!

"Best way no get hit: no be there!"
posted by BeerGrin at 12:59 PM on December 27, 2005

You can also turn the stats around on the sister and ask her why she's advocating that your GF gets a vehicle that has a higher than average chance of killing someone else.
posted by Good Brain at 1:12 PM on December 27, 2005

Independent of the ownership demographics, vehicles have inherent dynamic 'personalities' that are a function of power, weight, weight distribution, drivetrain, brakes, etc.

Safety is a measure of the driver's ability to understand the personality of the vehicle, and control it accordingly. [google video]

In terms of docile road manners in the widest range of conditions, I feel that an all-wheel-drive chassis is the clear choice. AWD is less prone to unexpected gymnastics (understeer or oversteer), and will go exactly where you point it up until it becomes unstuck entirely. You will pay a penalty in weight, fuel economy and repair costs for AWD. But for driving your way out of an accident, there is nothing better.

On preview: What Beergrin said Mr. Miyagi said.
posted by Triode at 1:19 PM on December 27, 2005

From personal experience, I'd also note that there's definitely more of a sense of detachment when driving a larger vehicle like an SUV. Some newer, smaller models have made strides in steering and suspension, but you'll still occasionally feel like you're steering around a large cardboard box when compared to a car.

Also, note that in any accident, it is likely that the SUV is the vehicle that causes more damage than it sustains, especially around smaller cars. The bumper is at hood/trunk level!
posted by mikeh at 1:23 PM on December 27, 2005

In general, it is true that larger cars will sustain less damage (and deliver a greater impact to the other vehicle) than smaller cars. However, this isn't always the case. For example, Fifth Gear's segment on the Smart showed that the diminutive car not only fared better than a regular compact coupe driving into a concrete barrier at 70 mph (!), but suffered less critical damage than a Mercedes S-Class in an offset head-on crash. The images of a roughly intact Smart taking a huge chunk out of an S-Class should be enough to convince you that bigger does not always equal safer even in the event of a crash.
posted by chrominance at 2:13 PM on December 27, 2005

But also note that all those people, who may or may not be injured, in all those cars that your girlfriend SUV just smacked into will be suing her, maybe past the limits of her insurance. Especially if it's their families suing her for the wrongful death of people in the other cars. This could effectively destroy your girlfriend's future.

Get a car with good crash ratings, lots of airbags, and good collision-avoidance, not something big and heavy. If she wants to spend extra on safety, get the safety options if there are any, or take a defensive-driving course. If she wants to be really safe, get a racing seat with a 5-point harness and a crash helmet.
posted by ROU_Xenophobe at 2:18 PM on December 27, 2005

Risk Compensation is the idea that people who perceive their situation as safer (whether or not physics agrees) will behave less cautiously in response. An archetypal demonstration of its affects occurred in the UK, where the number of driver deaths after the introduction of compulsory seatbelt legislation decreased, while the number of pedestrians and cyclists injured and killed increased, resulting in little overall change in death and injury rates in the population as a whole.

The popularity of SUVs, in part due to their perceived safety benefits, and despite their overall unimpressive safety record, is a further demonstration of risk compensation in action. Like everyone else said, it's the driver, not the vehicle, that has by far the most influence on safety.
posted by normy at 2:51 PM on December 27, 2005

ROU_Xenophobe: "If she wants to be really safe, get a racing seat with a 5-point harness and a crash helmet."

I assume you meant for her, because if it's for him while he rides with her, I foresee domestic unhappiness and lack of sexing.
posted by kcm at 3:40 PM on December 27, 2005

Volvo has a reputation for being an exceptionally safe car. Also, get lots of airbags.
posted by Ken McE at 4:32 PM on December 27, 2005

Anecdote: Recently saw a head-on crash between a small-ish Ford Focus and a Toyota Camry. The Camry was fucked up -- both air bags deployed, front-end destroyed. The Focus was merely dented. Do the reading from the advice above -- from just looking at the cars, their actual crash results are often counter-intuitive.
posted by frogan at 7:30 PM on December 27, 2005

After your buy that used car, go to tirerack.com and learn a bit about the tires it came with. They could be complete crap, especially if they're brand new.

Good tires will make any car safer. I really like Bridgestone S-03s. In the San Francisco Bay Area, they're great year-round. Very nice on wet roads. But don't use them in snow or temps below freezing.
posted by ryanrs at 10:55 PM on December 27, 2005

OK, it is seriously unwise to believe that degree of crush equals occupant protection, as several contributors have done. The Smart has only a tiny crumple zone by design; other Mercedes vehicles have large crumple zones. Do not make the macho mistake of equating a kind of human toughness (“That cannonball didn’t even leave a scratch on him”) with what happens in energy absorption in cars. The Focus or Smart occupants might be severely injured or dead, while the Mercedes and Camry occupants walked away. How bad the car looks after a crash is not an indicator of survivability.

The question the original poster should be asking is: What cars have better active safety or crash avoidance? Any car from a reputable manufacturer (and most in the U.S. are) has been well engineered for crash avoidance. However, some cars are better than others. I think a lot of European cars are explicitly overengineered for this, having read of many terrified auto journalists being taken for rides with e.g. BMW engineers who go from 100 mph to zero with hands off the steering wheel.

All-wheel drive and disc brakes (and, in the winter, winter tires – do not skimp on those) are an unbeatable combination, but I also would insist on traction control. Even cars that aren't overengineered like BMWs, but have all those features (some Subarus?), will be seriously manœuvrable and increase your chances of avoiding an accident.

All of this explicitly rules out nearly every SUV, with the arguable exception of the Volvo.
posted by joeclark at 6:16 AM on December 28, 2005 [1 favorite]

I second everything what joeclark said about tires and those crashes. Both the S-class and the Smart have very strong passenger safety cells but the S-class also has lots of metal in front of it that is sacrificed to absorb the energy of the crash and yes winter tires are a must in the winter. If she is buying a newish car, she should definitely get a car with one of those new-fangled stability control systems. They greatly reduce the risk of high-speed loss of control which is a common cause for the most dangerous accidents.

When buying an used car the most important thing for safety is the condition of the car. A cracked windshield or a steering that pulls or is numb in the middle can easily cause a big accident as can a lot of other thigns like poor tires or unbalanced brakes The safest car would be one that combines passive and active safety so it does not crash as often and survives crashes better. EuroNCAP has been crash-testing new cars since 1997 and new SUVS since 2002 and it shows that SUVs are not safer at all when not enjoying a weight advantage.
posted by fred_ashmore at 6:50 AM on December 28, 2005

Consumerreports.org on electronic stability control.
posted by fred_ashmore at 6:57 AM on December 28, 2005

from a mefite who's currently timeoutted:

An important note: Remember that a visual judgment of how "fucked up" a car looks after a crash is in no way indicative of its safety; there may be an inverse correlation between the after-crash integrity of the body and the damage done to the occupants for one reason: the car, and all occupants, have to stop. Imagine a car made out of megatitanium-7, an unbreakable alloy, smashing into a
wall at 60mph. The megatitanium body does not crumple or bend, so the total stopping distance of the car and the passenger is going to be miniscule, which is bad: the occupant will go from 60-0 in maybe .1 second instead of .2-.5 seconds. These are not accurate figures, but needless to say you want a high delta-t when you're talking about deceleration.

A car that gets "fucked up" uses crumple zones to increase the amount of time before the driver hits the seatbelt, airbag, or steering wheel. As long as the interior section remains intact, and I'm not suddenly smashed into a broken little sardine, I can tell you without a doubt that I'd prefer to be in a car with the crumple zones. I strongly recommend you keep this in mind when looking at vehicle options.
posted by amberglow at 2:03 PM on December 28, 2005

(he makes a really good point--you want the car to take the impact, not the people inside the car)
posted by amberglow at 2:12 PM on December 28, 2005