The study also found that the Chevy Cobalt and Saturn Ion were more likely to be involved in fatal crashes than their competitors.

Reuters, today, announced the results of an investigation into the General Motors ignition switch recall, claiming that at least 74 people have been killed in accidents with "similarities" to the accidents that GM attributed to the deaths of 13 people. Reuters is also claiming that the crash statistics analyzed, for the Chevrolet Cobalt and Saturn Ion, revealed that drivers were more likely to be involved in a fatal crash in the GM products than in competitive vehicles.

According to Reuters, it analyzed the incidence of single-car, frontal crash tests where the airbags did not deploy and at least one of the front-seat occupants was killed. It discovered that for every 100,000 Ions on the road, 5.6 were involved in fatal crashes. The Cobalt, meanwhile, had 4.1 fatal crashes for every 100,000 cars, while the Ford Focus, Honda Civic and Toyota Corolla had 2.9, 1.6 and 1.0 fatal crash per 100k vehicles.

Of course, you may be wondering how Reuters knows that 74 people were killed due to GM's ignition switch problems. The simple answer is, it doesn't, a fact that it acknowledges, writing:

"It is not clear how many of the deadly accidents identified by Reuters involved defective ignition switches, because crash reports typically do not include that data. That leaves open the possibility that air bags may have failed to deploy in some of the GM crashes for reasons other than faulty switches."

In fact, as pointed out by Insurance Institute for Highway Safety Executive Vice President and Chief Research Officer David Zuby to Reuters, it is entirely possible that "limitations" in the data analyzed may actually overstate the number of people killed due to undeployed airbags. One of the big "limitations" is the simple fact that sometimes airbags aren't meant to deploy, even in a frontal impact.

Then there's the issue of the statistical system Reuters used for its analysis, which isn't acknowledged in the story. Once again, it's the Fatality Analysis Reporting System, which is the same system responsible for an episode in March which saw Friedman Research Corporation claim that 303 people had been killed driving the Cobalt and Ion. Besides the fact that FARS doesn't, as Zuby points out, consider situations where airbags aren't meant to deploy, it also relies heavily on police data, which The Detroit News called "not always reliable," in its March report.

Those two factors are big reasons why FARS is only one of the statistical tools that the National Highway Traffic Safety Administration uses. The other is the National Automotive Sampling System/Crashworthiness Data System. According to our March story, which cited a 2009 IIHS study, NASS/CDS reported that airbags deployed in 45 percent of the fatal front accidents that FARS claimed lacked airbag deployment. We said it then, and we'll say it again now, FARS is simply not a reliable measure of airbag deployments.

This isn't meant to discount the Reuters' study, which Reuters itself points out may have resulted in an inflated figure, via its interview with Zuby. There is some value here, particularly in regards to the analysis of fatal crashes per 100,000 vehicles and how the Cobalt and Ion relate to the competition. Although it relies on FARS data, a similar analysis including NASS/CDS could provide a better picture of just how many people were killed due to the faulty ignition switches preventing airbag deployment.

Take a look below for an emailed statement from GM spokesman Greg Martin. Then head into Comments and let us know what you think of Reuters findings. Do you think the final death toll will be higher or lower?
Show full PR text
The total number of deaths associated with the ignition switch recall condition stands at 13.

The Fatalities Analysis Reporting System ("FARS") is a nationwide census providing NHTSA, Congress and the American public yearly data regarding fatal injuries suffered in motor vehicle traffic crashes. Although FARS contains information on fatalities within the recall population, GM does not have information on most of these crashes. GM arrived at the figure of 13 fatalities by assessing the detailed information in the claims data available to us. The criteria we used to make the determination was based on engineering expertise in both air bag deployment and electrical systems.

Airbags are supplemental restraints designed to deploy only in certain crashes. There are a number of factors that may contribute to an airbag deployment, such as severity of the cash, the angle of impact, occupant position and duration of the crash.

GM continues to monitor all claims we are made aware of in the recall population


I'm reporting this comment as:

Reported comments and users are reviewed by Autoblog staff 24 hours a day, seven days a week to determine whether they violate Community Guideline. Accounts are penalized for Community Guidelines violations and serious or repeated violations can lead to account termination.


    • 1 Second Ago
  • 29 Comments
      J
      • 6 Months Ago
      Why the heck is this company not closed yet?
      Larry Litmanen
      • 6 Months Ago
      Close that company already. The scary part is that WSJ reported that GM sales are up in month of May like 13%. People NEVER learn.
      lrx301
      • 6 Months Ago
      The simple truth is that these two GM cars, Ion and Cobalt, are not safe to drive. They have several times fatal crash rate than the Japanese counterparts.
        Car Guy
        • 6 Months Ago
        @lrx301
        Statistically you have a much high percentage of dying in a random car crash then by these defects.
          RocketRed
          • 6 Months Ago
          @Car Guy
          But the data point he is talking about is not limited to "these" defects, whatever they are. So I think it remains a valid point that it appears that the Cobalt has rather shocking difference in fatality incidence to comparable cars.
      Doug
      • 6 Months Ago
      So, the article title should really be that we don't have exact data, but we at Reuters are going to report something sensational and have you all believe us until we are proven wrong and do not issue a retraction at a later date.
        ChrisH
        • 6 Months Ago
        @Doug
        well its not like they used rocket motors or such. Still GM is getting away with a far more grievous transgression than Toyota did and paying hardly anything at all. Tell me the system is fair, fortunately the court of public opinion may fix that discrepancy.
          express2day
          • 6 Months Ago
          @ChrisH
          How is GM getting away with far more than Toyota did? GM is still very much in the midst of things and it's too early to tell how things will turn out, for better or worse. GM's $35 million fine was more than twice what Toyota paid and it's only the beginning.
        mapoftazifosho
        • 6 Months Ago
        @Doug
        I believe at this point the onus should be on GM to disprove findings like this, because it's not like they're going to come out with their own revised higher numbers. GM will do everything in their power to make this go away and another Auto Safety group believes the number could be MUCH higher...something like 300 deaths in these vehicles where the airbags failed to deploy. The problem with all numbers is that in all-likelihood these crashed vehicles have long since been destroyed and obtaining the additional evidence to prove or disprove is likely lost completely... I don't care what automaker this is...their feet should be held to the fire on this. I find it funny that most of the commentors on here will be more outraged with the headline than the fact that the number could be as high as 300 deaths, but no one will ever truly know... So
          tinted up
          • 6 Months Ago
          @mapoftazifosho
          Guilty until proven innocent? What a great way to run a justice system :/
          Patriot
          • 6 Months Ago
          @mapoftazifosho
          Because it is an attempt at grabbing headlines and trying to appear relevant. There is ZERO proof backing your statements. Absolutely nothing. My Daughter was hit head on in her 2006 Cobalt. The air bags deployed and she walked away while the other driver went to jail. Maybe your 300+ were drunk and unbuckled like 99.9% of the rest of these cases whether it be 3 or 3000. Typical person commenting without a clue.
          mapoftazifosho
          • 6 Months Ago
          @mapoftazifosho
          It's goddamn clear they're guilty of wrongdoing. At this point it's a matter of determining just how many people died as a matter of their carelessness...and I'm fairly certain GM would love for the death toll to stay at 13...but we all know that it will likely go no where but up. NHTSA Acting Administrator David Friedman told Reuters: "The final death toll associated with this safety defect is not known to NHTSA, but we believe it's likely that more than 13 lives were lost."
      krusshall
      • 6 Months Ago
      So your headline is absolutely misleading. Love the media.
      express2day
      • 6 Months Ago
      Anyone know what the average age of Cobalt or Ion drivers is versus Focus, Civic, and Corolla? If younger that could explain some, certainly not all, of the higher fatality rates for the Cobalt and Ion as younger drivers tend to be involved in more fatal crashes due to speed, less experience, etc.
        Larry Litmanen
        • 6 Months Ago
        @express2day
        I know that typical Corolla buyer is a middle aged female, so you are on to something here. At the same time we do not need any further proof that GM builds unsafe cars.
      Car Guy
      • 6 Months Ago
      I don't believe in karma but one has to wonder when GM screwed so many dealers and bondholders during their bankruptcy its now coming back to cost them dearly.
        David MacGillis
        • 6 Months Ago
        @Car Guy
        So what you are saying is that if those dealers had no product to sell they'd have been less screwed than they got in GM's bankruptcy?
      Ken Mam
      • 6 Months Ago
      Very biased-sided, more headline grabbing news article. In other news, Honda is recalling the civic for crappy cylinder gaskets resulting in engine failures.
      Bernard
      • 6 Months Ago
      And the 74 Darwin awards go too...
      normc32
      • 6 Months Ago
      Very nice journalism piece! Keep the great writing.
      icemilkcoffee
      • 6 Months Ago
      As an aside- why is the fatality rate so much higher for the Cobalt and Focus, than for the Civic and Corolla? Sounds like there are some serious deficiencies in the domestic cars.
        MZR2.3
        • 6 Months Ago
        @icemilkcoffee
        they tend to be made with union labor
        mbukukanyau
        • 6 Months Ago
        @icemilkcoffee
        They tend to be driven faster..
        Wm
        • 6 Months Ago
        @icemilkcoffee
        Death rates and crash tests are only partially linked. While Crash test are a controlled even test of the crashworthiness, death rates take into account the demographics of the drivers. GM, and to a lesser degree Ford, are generally preferred by less educated and less affluent drivers. GM in particular is becoming the trailer park queen. Late last year GM reported that 88% if their consumer receivables were sub-prime. They also said they would not longer provide that information since they prefer to hide the stigma for obvious reasons. Some cars with very good crash ratings have low death rates, while some with low crash scores have high death rates. In general it he call appeals adolescent males, it will most likely have a higher death rate. The highest death rate I've seen was for a 2dr, 2wd, S10 Blazer. Think about he people you've seen in those.
      RocketRed
      • 6 Months Ago
      It seems that Reuters is not actually saying what the title of this post says Reuters is saying, which is noted in the post. I know editors do the titles sometimes after the fact, but this is not good form. The most interesting bit of data here is the overall fatality rate of the comparable cars. I don't now if this controls for demographics and region etc., but it's hard to imagine other variables accounting for the fact that the Cobalt has 4X the rate of fatal accidents as a Corolla, a car that is ubiquitous across the country and bought by people in all walks of life.
        Bruce Lee
        • 6 Months Ago
        @RocketRed
        IIHS also regularly publishes fatality data by model, which I've found is helpful in figuring out what does better in the real world versus just crash test data. Crash tests are rated by size class so you tend to forget that smaller vehicles are actually much more dangerous in the real world. Too bad the latest data from IIHS was from 2011: http://www.iihs.org/externaldata/srdata/docs/sr4605.pdf When limited to the 2005-2008 model years it looks like the Cobalt had about 200% the fatality rate of the Corolla or Civic or Focus. The Civic probably had a higher fatality rate if you take into account more years because older models (pre-2001) did pretty badly in crashes. Nonetheless, the Cobalt had the worst fatality rates in it's class, and for that matter the Chevy Malibu also had the worst rates in it's class. Standing out as surprising to me was that the Subaru Legacy had the 2nd highest fatality rate after the Malibu in it's class. The older Subaru Legacies didn't score nearly as well in crash tests as their modern counterparts but nonetheless it still surprised me to see them 2nd-though it's possible that people are more likely to drive a Subaru in areas with bad weather conditions that also make it easier to get in accidents, and AWD may give people false confidence--I knew a lot of idiots who would insist to me that they "didn't need snow tires" because they had a Subaru and ignore the fact that braking speed and stability in the snow/ice are still entirely dependent on your tires. Either way I like to use a combination of crash test data and real-world results when figuring out which car would be a safe choice, and for whatever reason there's something about the Cobalt and Malibu that gave them the highest fatality rates. For that matter a Buick has the worst fatality rates in the study for large cars and the Aveo had the worst fatality rates in the mini class so real world data doesn't really do GM any favors. Conversely, in real world data Hondas and Toyotas and Fords seem to have better than average fatality rates. Anyways, it would seem that if you're paranoid about your teenager killing themselves it's best to let them drive a Ford Edge or Toyota Sienna lol.
    • Load More Comments