By on August 2, 2014



The Insurance Institute for Highway Safety has released the results of its latest round of small offset crash tests. This latest group of twelve cars posted a wide range of scores, highlighting the challenging nature of the Institute’s newest test. Only one car earned a “Good” rating from the Institute for this test, with several receiving the lowest score of “Poor.”

The complete results of the test are available here in convenient table format. Highlights include the Mini Cooper Countryman, which earned the only “Good” rating out of all the vehicles tested and retained its “Top Safety Pick” designation. Five vehicles earned acceptable ratings, including the Chevrolet Volt and Ford C-Max. Four vehicles earned a score of “Poor:” the Nissan Juke, Nissan Leaf, Fiat 500L, and Mazda 5. The Mazda 5’s test in particular is rather ghastly: the front passenger space experiences severe deformation, the driver’s door unlatches, and most worryingly, the driver’s curtain airbag completely fails to deploy. This performance, combined with a “Marginal” rating in the Institute’s side-impact test (making it the only 2014 model car to earn less than an “Acceptable” score), is a major black mark against the little MPV. America’s best-selling Serbian import was little better, exhibiting a great degree of passenger compartment deformation when crashed as seen in the photo above. Although it earned a “Poor” rating, IIHS pointed out that the Nissan Leaf experienced no battery discharge or other leakage when subjected to this fairly severe test; neither did the Chevrolet Volt. Though it earned an “Acceptable” rating in this test and not the highest score of “Good,” the Volt retains its “Top Safety Pick Plus” designation by virtue of its available front crash avoidance system.

The small overlap test, introduced in 2012, is an updated version of the classic “overlap” tests conducted by the Institute for many years. The old overlap test collided half of the front surface area of a car traveling 40 miles an hour with a solid barrier; the new test only allows a quarter of the car’s front area to be matched against the barrier. In other words, less of the car’s structure is made to take more of the force. The result is a greater chance of vehicle deformation, which puts occupants at greater risk of injury. The IIHS justified the new test on the grounds that such small offset crashes are responsible for a disproportionate amount of deaths and injuries in front-end collisions.

Get the latest TTAC e-Newsletter!

48 Comments on “New Round of IIHS Small Offset Tests a Mixed Bag...”

  • avatar
    SCE to AUX

    “less of the car’s structure is made to take more of the force”

    Actually, the force is the same – just distributed over a smaller area.

    Too bad for the Leaf, even though this is a very legitimate test. The Leaf could probably be improved, since it has a lot of empty space in this test’s crush zone.

    Guess I’ll just have to avoid drunks crossing over the yellow line. :(

    • 0 avatar

      Unfortunately if the dunk doesn’t get you the texting driver will.

    • 0 avatar

      Isn’t that saying the same thing? o_o

      I’d imagine two cars side swiping each other would be a lot different than hitting a stationary wall. I think if the other car has a good overlap crash rating, it will just kind of slide off of your car. So both cars will be safer.

      • 0 avatar

        Exactly what I was wondering. Wouldn’t the crumple of the other car result in less damage than this lobe of reinforced Berlin Wallite they use?

      • 0 avatar

        Not necessarily. The Wall isn’t moving, but another car moving in the opposite direction would add force by adding mass times velocity.

        The caveat is that unless both drivers freeze like the proverbial deer in the headlights, there’s likely to be some evasive action taken that can alter the angle of impact, for either better or worse. The test is, and must be, static for comparison’s sake, but real life situations have a plethora of complications.

        What surprises me is that the first test in 2012 revealed two strategies: beef up the passenger compartment, or redesign the front end/wheel geometry/suspension to deflect the force. The cars with poor results haven’t incorporated either one yet, and a door popping open and air bag not deploying are inexcusable.

        For low margin small cars, the makers may be putting off changes until the next redesign. That may be due to a lack of trust in the test, or the high cost of incorporating changes. I’ll bet on the latter, but I have to wonder if auto makers are growing weary of making expensive major redesign changes every time IIHS comes up with a new test.

        • 0 avatar


        • 0 avatar

          “but another car moving in the opposite direction would add force by adding mass times velocity.”

          Actually, force is mass times acceleration:


          What you describe is momentum, or mass times velocity:

          p = m*v

          A 3000 pound car moving at a steady 40 mph, hitting a stationary (but deformable) wall can be considered the same as a 3000 pound car moving at a steady 20 mph hitting an *identical* car also moving at 20 mph in the opposite direction.

          Agreed, the real world is more complicated, but the tests give good enough relative information.

          • 0 avatar
            Timur Apakidze

            The two scenarios you describe above are not the same.

            The kinetic energy of the moving car is absorbed by the structure of the the car. In case of a single car running into a barrier at 40 mph has 218 kJ of energy that the structure of the car needs to absorb by deforming in a plastic manner. To put that in perspective, this is equivalent to 2 ounces of TNT.

            For 2 identical cars running into each other at 20 mph, the total energy that the two cars need to absorb is only 109 kJ of energy. Not only that, if the collision is perfectly symmetric, the energy that either car needs to absorb is only 54.5 kJ per car.

            Remember, momentum is a vector, treating it like a scalar can lead to wrong results. Energy on the other hand is a scalar.

            — Your friendly physics nazi

          • 0 avatar

            >> The two scenarios you describe above are not the same.

            Ah, I see. Going from 40 mph to 0 mph is different (more kinetic energy) from going from 20 mph to 0 mph.

            But in a closed system, isn’t the energy involved with a car moving at 40 mph hitting a car at 0 mph the same as two cars going toward each other at 20 mph?

          • 0 avatar

            A car hitting a non-deformable barrier at 40 mph is roughly equivelent to it hitting an identical car at 80 mph combined closure speed (both cars going 40, or one car parked and the other going 80)

          • 0 avatar

            >> A car hitting a non-deformable barrier at 40 mph is roughly equivelent to it hitting an identical car at 80 mph combined closure speed

            Just to clarify, non-deformable is the keyword in this example. @Carve is citing Newton’s 3rd law when one body exerts a force an a 2nd body, the 2nd body exerts an equal and opposite reaction. Hence, 40 mph becomes 80 mph.

            For full frontal collision testing, the IIHS using a honey-combed deformable barrier. For the small overlap test, the barrier is as non-deformable as possible.

        • 0 avatar

          “unless both drivers freeze like the proverbial deer in the headlights, there’s likely to be some evasive action taken”
          That is the point of the small offset test. Direct head-on collisions rarely happen, because the driver(s) swerve. If they are lucky, they avoid each other entirely, and no collision means no statistic. If not, you get an offset collision.

    • 0 avatar

      The majority of fatalities occur in single vehicle accidents. Still, if someone comes to get you, it’s better to have a safer car.

  • avatar

    The underpinnings of the Mazda5 are really showing their age.

  • avatar
    The dummy was surprised.

  • avatar

    id be more worried about a half overlap, head-on, t-bone, or rear-ender than hitting an immovable barrier head on with a small portion of the front of my car, dead straight.

    its the law of diminishing returns here. most cars are so safe they have to come up with new tests for them to “fail”. its stuff like this that lead to gunslit windows and A-pillars that obstruct vision

    but hey- youre safe if you hit the edge of a concrete bridge embankment juuust right!

    • 0 avatar

      In the real world, the small overlap test serves as a decent proxy for those who steer to avoid hitting a tree or post, but fail.

      Almost succeeding with the maneuver but failing is worse than hitting it dead on because the benefit of the crush space is lost.

    • 0 avatar

      I drive a lot of 2 lanes, and it’s not as unusual as I would like to see two flatbeds loaded with the results of a small offset head on. Drunk, texting, or just trying to peak around an SUV at the beginning of a dotted passing permitted section; I don’t know all the causes. The heavier vehicle doesn’t always supply the least passenger area deformation, I like this test.

  • avatar

    Good lord .

    Being in The Auto Trade for most of my life I’m keenly aware of just how dangerous my oldies are , only my 30 year old Mercedes is any kind of safe but life’s a series of daily chances so I’ll keep driving my death traps .


  • avatar

    Also rated “Poor” in the small offset crash tests:
    Model T Touring Car
    BMW 600
    VW Bus
    Harley-Davidson CVO Softail Deluxe

    Two vehicles received “Exemplary” ratings:
    Caterpillar 777C mine hauler
    Baldwin 2-8-8-4 “Yellowstone” steam locomotive

    • 0 avatar


      Truly, refining these tests to Aspberger levels of fastidiousness will soon include impact from simulated meteor strikes.

      “While the Accord performed well in the Iron, Stony and Stony-Iron tests, it received only a Marginal for simulated Lunar and Martian rocks.”

      • 0 avatar

        The IIHS admits the small overlap frontal crashes are rare in the overall scheme of things. But when they do happen, they are among the most costly for insurance companies. Hence, they pushed this test on manufacturers in the hopes of reducing their own expenses. Money talks, and in this case, it’s a win-win for all those involved.

        • 0 avatar

          Except at what point does the cost of making things better in a very rare occurrence overcome the cost is terms of initial purchase and ability to see out of the car?

          I don’t buy cars to crash them, thus crash test results are at the absolute bottom of my list of priorities.

          • 0 avatar
            30-mile fetch

            Yes, it seems like we are going to hit a point of diminishing returns. Even if they can engineer a passenger cabin that doesn’t crumple at all you still have to deal with acceleration forces on the human body.

            I think too much emphasis is placed on some of these test scores; if you’re traveling another 5-10mph or the oncoming vehicle is heavier, the results go out the window.

          • 0 avatar

            It’s astounding that anyone goes to the trouble of studying medicine or building hospitals when virtually nobody wants to get sick. Fools, all of them.

          • 0 avatar

            I don’t care much about them too. Sure, higher all-around crash test ratings can determine how safe a car is but let’s face it – the tests are standardized and the manufacturers design the car in a way that would maximize the scores in these particular scenarios. But modify the parameters and the results go out of the window. The rule of thumb is generally that heavier means better (there are exceptions, of course); Americans have heeded that a bit too much though.

            I do not, however, understand why there are many people whose main reason for changing their car is the supposed higher safety. If you are so worried that your old car is so unsafe, why are you buying the new one? In about 7 years, that one would be considered unsafe too. Why would you EVER buy a car if the same, identical one would be considered safe at one point in time and a deathtrap at another, even if it would not have changed even a bit. I don’t get that but I guess every reason is good if you have your sights set on spending money.

            My current ride, the S40, has definitely been in an accident. There are sure signs of it both in the front and in the back, some panels do not align properly; the whole car has been repainted too. I truly don’t know if there are any airbags in the car. Maybe there are simply resistors simulating them installed in there. Maybe the covers are fake. I just don’t know. And frankly, I don’t care. I just try to prevent any situation in which I would find it out.

          • 0 avatar
            Chris FOM

            I don’t buy cars to crash them either but I’m not perfect and I do make mistakes. Also, other times someone else crashes my car for me.

          • 0 avatar

            Then you’re intentionally blinding yourself to information that might suddenly matter very much to you someday. And crash safety is something you just can’t tell from a look, a drive or a kick to the tires. I’m glad that all these tests are done. I can’t afford to have them done for myself. What’s wrong with knowing which of two competing cars is stronger underneath the skin?

          • 0 avatar


            Because ultimately, it doesn’t matter. Lady Luck plays a FAR greater role in the real world than whether car A did better on some artificial test than car B. In my personal case, any new car that meets the criteria that ARE important to me are guaranteed to be more than safe enough, and in most cases have been some of the safest cars available as determined by research into actual real world crashes.

            In the case of my classic cars, well, I know what I am getting into, literally and figuratively every time I get in them, and I drive accordingly.

            If you are driving responsibly, the chances of getting into an accident severe enough to badly injure you are actually very, very, very low. Bad things DO happen to nice people, but not very often!

  • avatar

    main thing I take from this round of these (not from this article since it didn’t warrant a mention or a tag) is how well the Lancer did considering its age

  • avatar

    I basically think of both the 5 and the CX-9 as being replacements to the MPV. But hey, poor crash test scores run in the family. The CX-9 also got a “Poor” in this test, and the MPV, well, go to the IIHS website and see.

    Hint: There weren’t any curtain airbags. At least not in the 2003 my parents had while I was growing up.

  • avatar


    Drive directly at what you are about to hit.

    • 0 avatar
      30-mile fetch

      Unless it is a vehicle badly outweighing yours and traveling at considerable speed. Since these tests are dependent on vehicle mass, hitting a bigger vehicle head on will still demolish you.

      With all the new electronic driver aids hitting the market, what I’m waiting for is a system that analyzes whether the impact is unavoidable and then wrests the steering wheel from your hands and sends you into a full head-on with the object rather than a glancing blow. Wouldn’t that be a trip!

    • 0 avatar

      Take a lesson from the Titanic – had she hit that iceberg straight on, she would have gone on to a hopefully long and successful career.

  • avatar

    Will there be two-tier insurance? One rate for the steering wheel and the other for the self driving? I think the one that comes without the steering wheel & pedals will be for the masses.

  • avatar

    This test shouldn’t be this impossible, I’ve seen a few wrecked junkyard cars that’ve taken impacts right at the edge of the front, a small ChevyGMC truck and a Volvo 240.

    I can’t say that they were going 40mph into a solid wall, but I can say that the Volvo took the crash well with plenty of space to spare between the crumple and the passengers compartment, you cut some metal and it’d probably still drive too.

    What we need is more space between the front bumper and us.

  • avatar

    I think the notion that the older cars are less safe is generally true, but there are definitely exceptions. See: Volvo XC90, which was introduced in its current form in 2003, and which scores quite well on this test.

  • avatar

    I stumbeled on a great website a few years ago called informedforlife.

    It ATTEMPTS to take the raw data (acceleration forces on crash test dummies rather than arbitrary claims like “good”) and factors them in with the danger and frequency of certain types of crashes, and then adds in a factor for mass (mass is your friend in a multi-vehicle accident) and tries to give one score.

    The data is often a mixed bag though and when data is unavailable they use a place holder for the average vehicle in that class. It’s still pretty useful though for identifying cars that are particularly safe or unsafe.

  • avatar

    “The Mazda 5’s test in particular is rather ghastly: the front passenger space experiences severe deformation, the driver’s door unlatches, and most worryingly, the driver’s curtain airbag completely fails to deploy.”

    I’d be far more worried about passenger compartment deformation than airbag performance. Especially so for an airbag that’s designed primarily for protection from impacts that are perpendicular to the applicable impact.

Read all comments

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

  • Skippity: Mustang II outsold S550. Double. In two less years.
  • nrd515: I hate almost everything about the Camaro, except it’s a great driver. The looks? Awful. The comfort...
  • Skippity: Not defending Nissan. But they issue a recall and some think it’s proof they make junk. Toyota has a recall...
  • la834: I don’t get why 2400w heat guns (or toasters, or dishwashers, etc.) aren’t readily available in...
  • 28-Cars-Later: I’m not sure what you mean, the 260 is unrelated to Our Lord of Eternal Torque.

New Car Research

Get a Free Dealer Quote

Who We Are

  • Adam Tonge
  • Bozi Tatarevic
  • Corey Lewis
  • Jo Borras
  • Mark Baruth
  • Ronnie Schreiber