Interpreting Crash Tests and Rollover Ratings

Overall, passenger vehicles are safer than they've ever been, but the crashworthiness of individual models varies greatly, even within a vehicle class. Because of this, understanding how a model is expected to perform in a crash is important before making your choice.

All vehicles sold in the United States must pass Federal Motor Vehicle Safety Standards, but the relative safety of all cars meeting this standard varies greatly.

The most important things to know about crash tests:

Not All Results Can Be Compared

A Ford Mustang is put through an IIHS crash test.

Model-to-model comparisons of frontal crash ratings are valid only within a vehicle class or between models of comparable weight (as long as they're within 250 pounds of each other). The test reflects how the vehicle would fare in a collision with another of the same model, not versus a larger or smaller vehicle (or a lower- or higher-riding vehicle). A heavier vehicle would protect its occupants better than a lighter one if all other factors were equal, but they never are. So a large vehicle with a Poor rating is not necessarily safer than a small vehicle with a Good rating. Unfortunately, researchers have not yet devised a reliable method for reporting the effect of size differences on a vehicle's score.

Note: Side-impact crash tests are comparable across classes because the sled that rams the test vehicles is of a consistent size and weight. See Side-Impact Ratings Have Deficiencies.

Likewise, the Insurance Institute for Highway Safety's rear-crash head-restraint ratings consistently test how well a stationary seat protects against whiplash by simulating a 20-mph rear crash. IIHS combines the results with an evaluation of the seat's geometry to arrive at a rating. Overall, the results are poor, but have improved markedly over the past year. Unlike the other tests, which also address chance of death, this test focuses on injury.

Back to top

Crash Tests Differ by Agency

The two testing agencies perform different types of frontal tests. The National Highway Traffic Safety Administration crashes cars head-on into a solid immovable barrier. Neither the angle nor the obstacle corresponds with the majority of real collisions. IIHS conducts a frontal-offset crash into a deformable barrier that reacts like another vehicle. This gauges how well half of the vehicle's front end absorbs crash energy. Many experts say this test is more revealing and better represents the majority of real-world crashes. Additionally, NHTSA has acknowledged that its New Car Assessment Program hasn't evolved quickly enough to account for marketwide improvements in vehicle crashworthiness. Simply put, too many models have high scores, and the differences among them — which do exist — aren't reflected. NHTSA is implementing updates that will first appear in 2010-model-year crash tests.

Back to top

Side-Impact Ratings Have Deficiencies

Side-impact crash tests currently aren't as telling. Though NHTSA has tested more models than IIHS for side-impact protection, we at Cars.com have come to dismiss these tests as inadequate for two reasons:

The sled employed to "T-bone" the stationary test vehicle has the height and mass of a car, not an SUV or a pickup truck. This tends to minimize its intrusion into the cabin — making it a best-case scenario.
NHTSA's chance-of-injury data are based on trauma to the test dummies' torsos, not their heads. Experience has shown that occupants' heads are more susceptible to injury in a side impact, and head injuries are more often serious and potentially fatal, according to IIHS.

A Mercedes-Benz E-Class undergoes an IIHS side-impact crash test.

NHTSA calls out a "safety concern" on some ratings on its website, but they don't affect the car's star rating. For example, the 2007 Chevrolet Cobalt two-door report stated: "Safety Concern: During the side-impact test, the head of the driver dummy struck the windowsill, causing a high head acceleration. Head impact events resulting in high accelerations have a higher likelihood of serious head trauma." Yet the car received a respectable double-four-star side-impact rating. That's misleading.

Tested with optional side curtain airbags, the 2007 Cobalt two-door report posted no safety concern but a three-star side-impact rating. Baffling. (Retested for 2008 with side curtains, now standard, the Cobalt scored four stars for side impact, and no Safety Concern was listed.) NHTSA will begin to factor head injury into its bottom-line results, but not until the 2010 model year. The IIHS side-impact test does measure head injury and employs a sled as high and heavy as a full-size SUV or pickup, a more dangerous scenario. Unfortunately, IIHS began this program recently, so ratings go back only a few model years. Because the sled is consistent, comparisons of side-impact ratings are valid across vehicle classes.

It's more important than ever to scrutinize crash-test reports — not just to determine if the car has side airbags, but if they are standard or optional on the model you're considering. In some cars, side bags have spelled the difference between a top and a bottom score, and it's up to you to make sure you get them. Side-impact tests use properly positioned, belted test dummies, which doesn't tell us what would happen if an occupant were out of position — in which case the side airbag itself can be hazardous, especially for children.

To address this, NHTSA reports now have an "SAB Out Of Position Testing" field that may read "Meets specifications." Know that this result has been reported by the automaker after voluntary testing — it's not performed by NHTSA. People concerned about injuries from side airbags should avoid the seat- or door-mounted type, or buy a car that disables them when the occupant is out of position. Honda and Acura have pioneered this feature. Curtain airbags are considered less dangerous.

Back to top

Rollover Ratings Have Shortcomings

Auto manufacturers and safety experts considered NHTSA's original Rollover Resistance Ratings, begun in the 2001 model year, inadequate at judging a model's rollover propensity because they were based solely on a mathematical calculation of the vehicle's center of gravity. Starting with the 2004 model year, NHTSA combined this calculation with a "fishhook" dynamic driving test in which the test vehicle swerves suddenly, then overcorrects. The combined results, called simply NHTSA Rollover Ratings, give a percentage chance of rollover — a star rating based on this chance and whether or not the model tipped up on two wheels during the fishhook test. While many see this as a step in the right direction, some automakers still criticize NHTSA for extrapolating some conclusions.

Back to top

Some Models are Not Rated

If the model you seek is missing, results may be pending or the vehicle may not be eligible. Both agencies concentrate on the highest-volume vehicles. Convertibles are rarely tested for this reason, though for the first time last year IIHS tested 10 models, including several best-sellers like the Chrysler Sebring and Ford Mustang. Results for new or recently reengineered models are likely to appear months after the car goes on sale because both agencies purchase their test subjects from dealerships. NHTSA notes if a vehicle is TBT (to be tested) or if results are pending or under review. IIHS has begun to offer more detailed information about whether or not test results are pending.

Back to top

Posted on 10/2/08