The Checkout

Consumer Reports Retracts Car Seat Study

Consumer Reports is retracting its car seat report after the National Highway Traffic Safety Administration said it had found flaws in its side-impact testing methods.

A quick recap: Consumer Reports tried to raise the bar on child safety seat testing by putting 12 popular car seats through 35 mph frontal and 38 mph side-impact crash tests. CR chose those speeds because they are the government standard for crash testing vehicles. Currently, federal standards require car seats be tested only in 30 mph frontal crash tests.

CR found that 10 seats failed and flew off their bases. One seat, the Evenflo Discovery, failed even at 30 mph. CR demanded NHTSA immediately recall it. Evenflo strongly disputed CR's findings.

After CR came out with its report earlier this month, NHTSA asked to review its data. Tuesday night and Wednesday morning, the agency apparently relayed its preliminary conclusions to CR, which promptly pulled the report pending further testing.

NHTSA administrator Nicole Nason said its initial review of the Consumer Reports testing procedures "showed a significant error in the manner in which it conducted and reported on its side-impact tests. The organization's data show its side-impact tests were actually conducted under conditions that would represent being struck in excess of 70 mph, twice as fast as the group claimed. When NHTSA tested the same child seats in conditions representing the 38.5 mph conditions claimed by Consumer Reports, the seats stayed in their bases as they should, instead of failing dramatically."

Consumer Reports put out a statement saying, "Our tests were intended to simulate side crashes at the NCAP speed of 38 mph. The new information raises a question about whether the tests accurately simulated that speed, however, so we are now reviewing our tests and the resulting article."

What does this mean for you and me?

Well, until this gets sorted out, both sides reiterate that it's always better to use a car seat than not to use one at all.

CR also asked consumers "to suspend judgment on the merits of individual products until the new testing has been completed and the report re-published."

CR had previously recommended consumers who have car seats that failed its testing to keep using them, but to attach them to the car with seat belts and not the LATCH system. We'll see if they stick by their guns on this one following the latest round of tests.

We'll let you know as soon as we do what the new results are. Until then, my question for you is: Does the retraction shake your faith in Consumer Reports?

By Annys Shin |  January 18, 2007; 3:00 PM ET Consumer Alerts
Previous: Anticipating Tax Refunds May Cost You | Next: You Qualify ... to Spend More


Please email us to report offensive comments.

No, it doesn't shake my faith in Consumer Reports - on the contrary, it's admirable that they're withdrawing their report instead of blindly proclaiming that they're right. That tells me that they're interested in using the scientific method and getting it right.

Posted by: h3 | January 18, 2007 3:57 PM


Told you so.

h3, you need to re-evaluate your position. CR put out that report, without double checking themselves. They have turned into a reckless organization. I'd agree with you if NHTSA wasn't the folks that issued the contrary report, and CR had done their own follow up and retracted voluntarily. Their magazine is rife with technical misunderstandings and mistakes.

Posted by: Kim | January 18, 2007 4:10 PM

My faith in Consumer Reports wasn't very high to begin with. This retraction occurred only after the government found significant errors in the testing including testing side impact crashes at twice the speed consumer reports said they tested them.
Bottom line on car seats remains --install them correctly and use them correctly all the time.

Posted by: downtown mom | January 18, 2007 4:11 PM

Whether or not consumer report's crash test methods correctly simulated a 35 mph side impact crash should not necessarily invalidate the results.

As long as the same method was used with all the car seats, the results still indicate that some seats and attachment methods are better than others, which is to me, valuable information. Not all accidents happen at government-approved speeds of 30-35 mph.

Posted by: mom2B | January 18, 2007 4:19 PM

While I agree with the previous poster that CR demonstrated their integrity by immediately pulling the report, I am gravely concerned with the apparent failure to properly vet test results prior to publication. Given the serious nature of the findings, I would have expected CR to have been absolutely certain of their validity. That it only required a few hours for the NTSB to raise serious questions - virtually a 100% discrepancy in regards to the test speed is not a trivial mistake - leads me to believe there was a critical lapse in process, judgment, or both at the CR lab. I am quite anxious to hear their response and explanation.

Posted by: Concerned | January 18, 2007 4:20 PM

I'm disgusted by this whole episode. The original report scared the heck out of me because I had one of the failing seats and was using it for my 4-week-old daughter. How could I in good conscience keep using such an inferior product when it would jeopardize the safety of my child?

When we called the manufacturer, they insisted there was nothing wrong with their product and that they would not be recalling the seats. So I joined the stampede of parents at the baby superstore and picked up one of the last remaining "Consumer Reports approved" carseats.

I was angry at the carseat manufacturers for not acknowledging this report, but I expected that reaction from them. I'd trusted Consumer Reports, and they've betrayed that trust and caused me and other parents like me all sorts of stress and unnecessary financial hardship. I wonder if the editors would like to buy one of these carseats back from me?

Posted by: Mike | January 18, 2007 4:34 PM

At least Consumer Reports is willing to admit to their error -- many manufacturers would not do so.

Posted by: Nate | January 18, 2007 4:43 PM

For the person unconcerned about the discrepancy between CR stating that tested the seats at 30-35 mph when in fact they tested at 70 mph: That really bothers me. That is a major mistake and it apparently went unnoticed. Testing correctly is important --how do we know that the "safe" seats are really "safe"?

In addition, to Nate many manufacturers have voluntarily pulled products when safety issues are found. Not all manufactures, sure, but many do.
I too will be very interested in how CR attempts to get out of this major snafu. I expect a few jobs will be lost there.

Posted by: downtown mom | January 18, 2007 4:53 PM

They're still the best thing around for independent product analysis - but this is a real shocker. My faith is shaken but not toppled.

Posted by: Dave B | January 18, 2007 5:21 PM


You're incorrect in your assertion that there's some useful information in CR's baby seat fiasco. If they're testing seats WAAAY beyond their design or mission requirements you can't really trust their data.

Is it fair to say that my car is unsafe because it can't withstand the heat of reentry from orbit? Of course not, Toyota didn't design cars with that in mind in 1997. A 70mph side impact test is hardly a realistic scenario.

Posted by: Kim | January 18, 2007 5:24 PM

It really comes down to:

Who watches the Watchmen?

Posted by: Rorschach | January 18, 2007 5:30 PM

Yeah, 70 mph impacts are clearly not realistic since the never happen...yeah, right. They screwed up, they were called on it, they retracted, get over it. This happens in research all the time; Sorry for you who were scarred for life. Somebody does a study, they submit it to a peer review process, and somebody finds a mistake. The best way for CR to have presented this is to have peer reviewed it prior to publication (which is what happens typically in research), but I'm pretty sure there isn't the infrastructure in place. I bet you money that if CR asked NHST to look at it prior to the big outcry, they wouldn't have, or if they would have agreed to it they would have taken their sweet time.

Posted by: Columbia | January 18, 2007 6:11 PM

When it comes to my son (who rides in a Britax Marathon) I'd rather have the rejection criteria set too high than too low. It's safer for our babies if we accidentally reject a few good ones, instead of setting the bar so low we keep the bad ones on the market.

Anyway, as the previous poster said, a crash at 70mph is more likely than at 30. My question is, why do the federal regs only test at 30 anyway? You want shaken faith, look there.

Posted by: jaxoms_mommy | January 18, 2007 6:36 PM

When it comes to my son (who rides in a Britax Marathon) I'd rather have the rejection criteria set too high than too low. It's safer for our babies if we accidentally reject a few good ones, instead of setting the bar so low we keep the bad ones on the market.

Anyway, as the previous poster said, a crash at 70mph is more likely than at 30. My question is, why do the federal regs only test at 30 anyway? You want shaken faith, look there.

Posted by: jaxoms_mommy | January 18, 2007 6:40 PM

CR remains the only independent, self-funded, no-studies-being-bought-by-manufacturers, testing organization around. I'll take their occasional mistake, especially when they've demonstrated a willingness to listen to contrary opinions. Cuz when they're right, they absolutely stick to it; Suzuki Samurai, anyone?

Posted by: jcm | January 18, 2007 6:40 PM

I still feel comfortable with Comsumer Reports. Even if in error, I see a willingness to acknowledge error. I'm not a parent who would've gone out to replace a failing seat, but expecting a baby the study did guide my choice in a new seat. Given the limited time these seats are in use, if I owned a failing one, I'd have simply moved up to a convertible carseat sooner. The report did make me aware of things I wasn't previously - like Britax makes a great infant seat boot I'd like to see offered here, and that testing is only for 30mph frontal. It seems well below the rigorousness I would hope for, and just isn't the level of accident that I carefully use car seats in dread of.

Posted by: expecting | January 18, 2007 7:56 PM

I hope the whole thing leads to the NHTSA moving to conduct side impact testing for impact seats as a standard. It's obviously worth addressing.

Posted by: dlocsaeci | January 18, 2007 8:06 PM

What would shake my faith in CR is if they had erred the other NOT being stringent enough in their testing and had declared a product satisfactory when, in fact, it was not. CR is an excellent organization BECAUSE it has "no dog in the fight." The steps that CR has announced to investigate itself are precisely the steps they should be taking...and if it turns out that their testing procedures were improper, I'll bet that they will acknowledge that fact immediately and move on.

Posted by: est | January 18, 2007 8:07 PM


You folks want to vilify the government agencies and manufacturers, but are willing to give CR a pass. Sad.

CR has become arrogant, and ignorant. They once championed the average consumer, but now they merely seem to exist for their own amusement.

A 70mph impact on the side of your stationary car. That happens every freekin' day, no? That's what CR was testing/simulating: A stationary victim car that was getting hammered by a 70mph car. Regardless of a child restraint, that's a tragedy.

Most accidents are dynamic, and a 30mph side impact, seems more reasonable to that effect. A 70mph head-on, or glancing blow (off angle),accident simulation may also be warranted (a collision with both vehicles are travelling @ 35mph).

But nobody lives in a 70mph side impact.

Posted by: Kim | January 18, 2007 8:53 PM

Kim, I am curious about your stance that Consumer Reports has become a an "arrogant," "ignorant," and "reckless" organization and that the magazine is "rife with technical misunderstandings and mistakes." I have not heard this before and I am curious about what makes you think this. Are there things other than this carseat report that have been problematic? I'd like to know about them!


Posted by: madcap150 | January 18, 2007 10:56 PM

I've used CR since I was a child for ratings and information on a world of products. I take this as a mistake... but, since CR is non-profit... no ads... they are really all we have for non-paid-for ratings. Taken on the whole, they are a whole lot more reliable than the "raters" who take ads to pay for their work.

Posted by: thewahoo76 | January 18, 2007 11:03 PM

some silly comments out there exposing a real lack of critical thinking. How about someone show me one 70 mph side impact accident in the past 12 months - its a near absurd situation that at best occurs in some action movie stunt scenes.

This was not just CR made a mistake they messed up on the basic data analysis - they were wrong at the root of what they do. That should call into question all their data analysis that has not been indpenedently verified.

Unlike some silly commenter said they are not part of a peer review process - they do not do studies and submit it for review. So the only way a mistake is caught is when it raises a bigest enough fuss to be tested. Think of all the monetary damage done by data errors that have not been corrected and just mistakingly labeled one product as failing or passing mistakingly.

They may be all we have for unbiased reviews but if their data analysis is unreliable than their reviews are completely worthless.

Posted by: aa | January 18, 2007 11:20 PM

Who do I trust, a nonprofit that lives and dies by it's reputation and corrects things when they get them wrong? Or a corporations whose stated singular goal is to make the maximum amount of money for shareholders?

I also think that the standard of 35 mph is ridiculous. If someone runs a red light, they probably are NOT going at 35 miles an hour and me and my child are just sitting ducks.

I think the government only tests at a level they think the products will pass. The standards are WAY WAY too low!

My faith in Consumer Reports is unshaken. My faith in government testing of corporate products is always going to be minimal at best.

Posted by: Shannon | January 19, 2007 12:15 AM

I have to say that I agree with CR supporters, to an extent. While I'm shaken by the knowledge that such a respected magazine would publish an article with such a glaring error, I would much rather purchase a car seat for my impending infant that meets and exceeds federal and independent testing guidelines than one that just meets the already low federal standards for crash safety.

The reality is that crashes do occur at higher speeds than 30/35mph, and these are the crashes that are seriously endangering the life of your child. I dont want to be driving down the road wondering if my baby is in the safest possible seat. I'm glad to know that organizations like CR are looking out for consumers and when they err, admit it and launch investigations to get to the bottom of things.

Posted by: Portland, OR | January 19, 2007 12:31 AM

Consumer Reports offers an Unbiased opinion and research. Everyone at CU knows the difference between 70 and 38. The issue here will end up being a question of testing methods. How many G's where generated and why. It is no where near as simple as you might want to believe it is. If all seats are subjected to the same test then the test is valid. If you are looking for a devil, consider this. We have two cars, a small car and a 9000 pound truck. When cars are tested in an impact for safety they are tested against similar weight vehicles. A sub compact that has a five star frontal crash rating will be destroyed by a full size pickup that fails the test and it is highly likely that the driver of the poorly rated vehicle will survive and the five star driver will not because the truck will stop the car in its tracks and throw it back in the direction that it came from. The truck will sustain substantially less impact force that the car at a rate similar to the ratio of the vehicles weight difference, in this case 4 to 1. All that is needed to see the value in these test is to see the difference between getting hit by a truck or another small car. If you choose not to believe this then explain why vehicle weight is consistently the most important factor in a collision. In my small car I always use the seatbelt to attach the seat to the car because the base is more likely to fail in any collision the small car has. Also you should consider that all of the seats manufactured will fail at some speed. Just like the way a child is suddenly safe to have small parts in their toys at three there is this arbitrary 30 mph that is used for testing. If two cars are going 55 mph and the hit head on and they are exactly the same weight, the occupants in each car will experience a 55 mph barrier collision. If one car weighs twice as much as the other it will have twice the kinetic energy at impact driving the other car backwards. The car that was traveling 55 and is lighter will experience a collision force greater than the speed at which they were traveling. What if you get hit by a driver that is exceeding the speed limit? The whole testing methods thing is nonsense. Seats should be tested until they fail, lessons should be learned and designs improved. Companies will continue to make seats that just meet the standard because they will make more money for their investors that way. The whole thing is one great big smoke and mirrors illusion where the consumer gets fleeced by the companies and the government until someone else, like CU in this case forces them all to raise the bar.

Posted by: SN Barrett | January 19, 2007 4:38 AM

SN Barett, great post.

Posted by: wg | January 19, 2007 9:23 AM

SN Barrett- VERY TRUE! I am all for improving vehicle standards. As I stated in the last article, manufacturers test for the cheapest materials they can get away with using. A solid vehicle will always fare better. If the car seat was tested at 70mph, I would rather have that data knowing that I want a device that can take more punishment than 38mph. The seats should be tested in the weakest vehicle on the road. Then people will see that vehicle composition plays a huge role in safety- something the manufacturers would have you forget.

Posted by: Chris | January 19, 2007 9:43 AM

My folks had CR when I was a kid. I went off, grew up, and became and engineer. Got married, realized I needed to buy things I really didn't know about, so I picked up a subsription to CR.

It's good for laughs, but not for analysis. And too many folks take what they say for gospel instead of an opinion, or start of an educated opinion. Also, a lot of thier information comes from subscribers, who take CR's word as gospel, thus feeding more faulty data into their process.

While it is valuable to have an independent organization to aid the consumer, we should not blindly accept their results, just as we should not blindly accept protections and assurances from the Government Agencies or manufacturers.

SN Barrett- you're partially right. The two autos in your 55 mph collision will feel the forces of the collision, the design of the autos will dictate how much of that is distributed to the occupants. If a car tranferred all of it's energy to me, I doubt I'd survive a 10 mph collision.

Posted by: Kim | January 19, 2007 10:04 AM

Having seen the results of a 70 mph side impact, I can say that no matter what baby seat your kid is in, they're dead unless they're on the contralateral side (center won't cut it). My friend's car was hit on the side at around 70 mph and had the passenger door pushed in almost past transmission. Thankfully, he wasn't hurt since he was the driver, but had anyone been sitting in the passenger seat, they would have been pulverized.

CR was irresponsible in their recommendations when they claimed that they were using federal standards. They should be criticized for misleading the public, just as any organization, public, private, or "in the public interest" should when it screws up.

Posted by: AK | January 19, 2007 10:54 AM

Having seen the results of a 70 mph side impact, I can say that no matter what baby seat your kid is in, they're dead unless they're on the contralateral side (center won't cut it). My friend's car was hit on the side at around 70 mph and had the passenger door pushed in almost past transmission. Thankfully, he wasn't hurt since he was the driver, but had anyone been sitting in the passenger seat, they would have been pulverized.

CR was irresponsible in their recommendations when they claimed that they were using federal standards. They should be criticized for misleading the public, just as any organization, public, private, or "in the public interest" should when it screws up.

Posted by: AK | January 19, 2007 10:56 AM

I am growing more and more skeptical of the comments I see posted for the various checkout columns and cannot help but wonder if corporate lobbyist/professional marketer types are not orchestrating organized posting campaigns.

The response to this particular column and the id theft column earlier this week have struck me as marking a major shift in commenting that is skewed toward those who would have a financial interest in downplaying identity theft and downplaying organizations such as consumer reports which are one of the few organizations that seem to try to act on behalf of ordinary folks.

Posted by: Skeptical of Posters | January 19, 2007 3:34 PM

I never had any faith in consumer reports.

Posted by: mf | January 19, 2007 9:08 PM

how many of us actually rode in car seats as a child? (perhaps I'm simply showing my age) even if a given seat failed a given test, is it actually more dangerous than no seat at all?

wrt auto safety and weight, does that mean that a '57 chevy is safer than a '05 Corolla?

no source is infallible.

Posted by: rick jones | January 20, 2007 1:49 PM

Maybe all the CR supporters should ask themselves where CR gets all the money to purchase all the products they test every year. Most people know that magazines and newspapers don't pay the bills with subscriptions let alone CR which has to spend millions to purchase the products they test. But I would be willing to bet that many a lawyer was licking their chops at the prospects of a class action suit filed on behalf of all those poor consumers who purchased seats from these evil manufacturers. And then ask yourself where all that unclaimed monies from the damages awards from these lawsuits goes when they offer the wronged consumer $3.75 that no one bothers to claim. Of course the attorneys have earned millions for their efforts in protecting the poor consumer, but little known is that the unclaimed monies are awarded to consumer organizations such as Consumer Union parent company of CR. So if you really want to know what motivates CR maybe you should follow the money.

Posted by: ES | January 20, 2007 7:03 PM

I have thought consumers report reviews of automobiles has been bias against domestic companies for years.

Posted by: SM | January 20, 2007 8:13 PM

Given that it was not worth the possibility of the car seat (for our new baby) to not be safe, we replaced our "unsafe" car seat with a new "safe" one. What gets me is that CR did not acknowledge the thousands of consumers who, like me, spent $100+ to heed their advice (out of fear). While I can now infer that my new seat is still safer than the old one--all I have is inference--I will be much less inclined to trust them in the future.

Posted by: jrm | January 20, 2007 10:53 PM

I'm going to jump in and defend CR too. I think it's great that they recalled their report and are investigating into it more. It could be an honest mistake. What I'd love them to do is come out with their report showing which car seats stood up in the "average" crash test of 35-38mph and which withstood the highway speeds of around 70mph that news articles are saying they mistakenly tested them at. A couple people posted about how rare 70mph side impacts are, no one lives, etc. Guess what folks-never say never. I've survived one, and my car seat sitting in the middle seat in the back (thank goodness without my child in it) wasn't even cracked. There are so many speeds that crashes occur at, different angles, etc. Why not test at the higher speeds? I feel for all the parents that went out and bought new safer car seats, but the bottom line is, how much is your child's life worth? And wouldn't you rather have a seat that surpasses average in safety tests? I still will trust CR, and unlike many government agencies they at least owned up to the mistake and are looking into it.

Posted by: abc | January 21, 2007 2:50 PM

I think CR is a fine organization. Everyone makes a mistake every now and then.

We happen to have the #1 car seat. I'm glad to see it works at 70 mph!

Posted by: new mom | January 21, 2007 5:23 PM

I have had an online subscription to CR for a couple of years. I use it to check on products I am going to buy. I really have no other choice when I am looking to replace my vacuum cleaner than to look at their review or buy 20 vacuum cleaners, test them all, then take back the 19 I didn't want. I also would like the poster who spouted off with the 'cr is arrogant, and ignorant' remark to back it up with something. That IS the point of CR, no? Backing up the research? FWIW, on the roads I drive it is dangerous to go less than 70 mph. I also have Washington 'Checkbook" magazine that comes a couple of times a year. Talk about 'testimonials'. I will continue to support CR with my measley few dollars a year for the subscription. I will continue to look at their research before I make a large purchase. I wonder if the people commenting on this are looking at other headlines in the news and wondering about arrogance, false and misleading reports and wasted resources. All that being said, I have to agree with ES--follow the money.
CR is still a good publication.

Posted by: jj | January 24, 2007 10:33 AM

I think the biggest point that we are all missing here is that regardless of an error or not....I'm sure glad I can see what car seats would fail based on the actual speed that I drive. Most of the major highways I travel on allow a speed of 70 mph. It is very important to me to know which seat holds up the best if I were to get into an accident while driving the speed limit. I am thankful for CR even if it were an error on their part. I'm somewhat appauled at the idea of these tests being conducted at only a certain speed vs. multiple speeds. GO CR!

Posted by: Sandra | February 7, 2007 12:01 AM

The comments to this entry are closed.


© 2010 The Washington Post Company