March 3, 2009

Whoa. Tuesday. Freakout. Car Seats Fly In Undisclosed Government Crash Tests


I'd say I hope you're sitting down when you read this, but there's probably an unpublished NHTSA crash test that shows your seat's a dangerous flying projectile, too.

An investigation by the nation's de facto consumer safety authority, the Chicago Tribune, found undisclosed government data that showed when they were subjected to actual car crash tests, otherwise-safety-certified infant car seats would often go flying and cause severe injuries not only to the baby dummy, but also to other passengers in the car.

As you may know--or probably not--car seats are not actually tested in cars, but on a crash sled. Over the years, researchers have clearly identified many ways in which the test and reality diverge: no side impact tests, no front seat for the car seat to slam into, etc.

But the car seat manufacturers consistently fight any attempts to strengthen safety standards by, say, testing car seats the way they test cars: in actual crashes. They say it'd be too confusing and expensive to certify car seats for specific models of car; better to have a single, universal standard. Yeah, well.

But as it also turns out, vehicle safety ratings do not take car seats--or any rear seat safety measures, for that matter--into account at all. The Tribune found 5-star safety-rated minivans, for example, where the rear seats would separate and sent a car seat flying. But it's OK; they weren't being tested on it.

Definitely read the Tribune story for the details about which models of car seats and cars did not fare so well, and to find out which company's general counsel rejected all the test results as "worthless" [hint: Graco], and which company was so shocked when they first learned of the tests in 2007, they switched to Canadian crash testing for their US car seats [hint: Combi].

Car seat tests reveal 'flaws' [ via dt reader gfr]


How do we get the Insurance Institute for Highway Safety in on this? They're already doing more real-world crash tests on vehicles than NHTSA does.

I would love to have a reliable third-party testing these seats, and as an independent not-for-profit, IIHS could start testing without the influence of lobbyists, provided it's membership (auto insurers) would support it.

Years ago when I took a carseat installation class - I was told that the snap-in bases for the infant carriers were not safety tested above 35mph and were considered fairly unreliable in higher speed crashes because they tended to disconnect upon high speed impacts. I went the Maxi-Cosi way and used the seat belt to lash my kid in. But I guess the larger convertible car seats would be safer.

I wish they had provided a list of the carseats that did well in this test, if any. If they did, it was buried somewhere else on the website. After reading this, my only question is which seat should I buy instead? They didn't provide an answer to that question.

I'm confused about what I'm seeing in those videos. I am most concerned about the way the seats fly up and turn over... Is there some sort of explanation of what should be happening? The video for the Chicco Key Fit looked similar to that of the Graco, yet they were rated differently.

My daughter spent 90% of her infant journeys in a Britax Diplomat convertible, yet I wonder how the installation of that is any safer at speeds in excess of 35MPH than any of the infant seats in that video.

Something needs to change in this industry.

One of the mfr complaints about these tests is that they're not systematic--the Graco lawyer called them "purely experimental"--so they don't generate a comprehensive system of ratings and rankings. Also, the results would vary by car model, so it was possible a seat did fine in one car, but would separate from the base in another. [The Graco stayed in the base 4 out of 5 times. Graco claimed the technicians installed the seat incorrectly the fifth time. NHTSA experts showed them that was not the case.]

Unless your carmaker maintains a list of recommended/tested/approved car seat models, it won't exist. Car seat mfrs will insist their seats are safe because "they pass all government required safety tests." Frankly, I can see how it'd be a PITA to maintain a giant matrix of car and car seat compatibility rankings. For any traction, there needs to be some incentive for mfrs to want to gain "approved/recommended by" status.

Yet another reason I'm glad we never bought a bucket for our kids. Convertibles from the start, because the actual SEAT is attached to the car. It just made more sense to me to have a big strap holing the seat in, rather tan a little plastic tab.

I sure hope people aren't getting hurt because of faulty testing. Scary.

When I first read this article I was astounded by the fact that the left hand refused to talk to the right hand, or, in this case, the NHTSA and the Consumer Products Safety Commission. You'd think that one gov't agency would offer information to, and the information would be acted upon, by another gov't agency. It's like the CIA and the FBI refusing to cooperate on busting criminals. Oh wait, they already do refuse to cooperate.

Anyways, it seems we need some greater accountability in these agencies. It seems they could save a lot of money if, instead of crashing a car twice (once for the NHTSA and once for the CPSC) they both shared the costs of crashing a single car. But then that might upset car seat manufacturers because the NHTSA would inevitably require more "real world" testing. And it is becoming increasingly obvious that the CPSC is beholden to manufacturers.

Car seats are a particular regulatory mess because they're covered by two agencies, one each for car and non-car uses. Never mind that the car seat working group for ASTM or whoever is all the same people. This is why there have been 10 million car seats recalled in the last ten years for handle failure; it's not a vehicular use, so it falls under CPSC. Meanwhile, everyone keeps thinking car seats should be totally safe wherever they use them because they're "certified".

Hey guys,
I blogged about this here for

Now, as a dad I of course find this really sketchy on a number of levels but note the statistics I found:
there were 95 fatalities of children under the age of one in 2007 out of over 40,000.

Every death is horrible, especially infants but somehow this number has become very small since the advent of the child safety seat. So they have to be doing something right.

As a car reviewer who puts a seat in a different car weekly, I think there's more to be said by wedging a seat in behind the front seats than anything else. Of course a lot of cars you can't do that with at all.

Also there was no description of these tests in depth. The NHTSA called them research.

The expense of running a separate test for every model just to get car seat results would be large as well. Some manufacturers do their own testing and tell buyers which seats work best in their cars. But that's not a 3rd party.

I think part of the problem is there are a lot of car seats on the market and a lot of cars. You couldn't expect a car manufacturer to build a car to take into account all those seats.

NHTSA has been horribly run the past 8 years by a woman who admitted she didn't install her own children's car seats properly.

Either a new standard needs to be devised or a 3rd party like IIHS has to do it. Remember when CR tried this they did it wrong!

You know, the Europeans are light years ahead of us on this. They routinely do side impact tests, for example, in Germany to give TUV and other certifications to carseats over there. That's why we buy our Maxi-Cosi in Europe, so we know our kids will be safe. They have tests like those that our government thinks are too expensive or too time consuming -- and they put the results out for the public to see. BMW, for example, had a section of their German website detailing which car seats actually work in their cars.

Google DT

Contact DT

Daddy Types is published by Greg Allen with the help of readers like you.
Got tips, advice, questions, and suggestions? Send them to:
greg [at] daddytypes [dot] com

Join the [eventual] Daddy Types mailing list!



copyright 2018 daddy types, llc.
no unauthorized commercial reuse.
privacy and terms of use
published using movable type