IIHS REPORT: The Risk of Death While Driving Different Vehicles
Many passenger vehicle drivers died during 2000-03, but the risk of death isn’t the same in one vehicle compared with another. Car, minivan, SUV, and pickup truck models vary widely in the likelihood of dying in a crash. The average driver death rate in 1999-2002 passenger vehicle models during 2000-03 was 87 per million registered vehicle years. But the death rates in some models were two or even three times as high (see tables), while the rates in other vehicles were much lower. Large cars and minivans dominate among vehicle models with very low death rates. The models with the highest rates are mostly small cars and small and midsize SUVs, many of which also have high rates of death in single vehicle rollover crashes. The model with the highest death rate of all — the two-door, two-wheel-drive Chevrolet Blazer with 308 driver deaths per million registered years — also had the highest rollover death rate (251 per million).
“Many of these general patterns of death rates have been consistent since the Institute began computing the rates by vehicle make and model in the late 1980s,” says Institute chief operating officer Adrian Lund (see Status Report, Nov. 25, 1989). “Since then there also has been a pattern of improvement. In the late 1980s the overall driver death rate was higher than 100. The latest overall rate was 87.”
Vehicle body style, size, and fatality risk: Important characteristics of vehicles that influence their driver death rates are type, body style, size, and weight. Within virtually every group of vehicles, the smaller and lighter models have the higher rates (see table). Among cars, for example, the smallest two door models had the highest death rate at 190 per million vehicle years. This rate is more than twice as high as the average for all vehicles included in the study.
Midsize sports cars also had a high rate at 133 driver deaths per million vehicle years. This was higher than for either small or mini sports cars, so this type of vehicle was an exception to the general rule that bigger means lower death rates.
The vehicle group with the lowest driver death rate was large luxury cars with 37 deaths per million vehicle years. The next lowest rate was in large minivans and station wagons with 42 deaths per million.
Vehicle weight and the risk of death: Because vehicle size and weight are so closely related, it shouldn’t be surprising that their effects on driver death rates are similar. In each group (cars, SUVs, pickups) the heavier vehicles, like bigger ones, generally had lower death rates (see table, p. 7). The rate in the lightest SUVs, for example, was more than twice as high as in the heaviest SUVs. “Pound for pound across the vehicle types, cars almost always have lower death rates than either pickups or SUVs. This generally is because the SUVs and pickups have much higher rates of death in single-vehicle rollover crashes,” Lund explains. In some weight groups, the death rates in cars were dramatically lower. For example, the rate in cars weighing 3,501 to 4,000 pounds was about half of the rates in pickups or SUVs of similar weight. The exception was light pick ups, which had relatively low rates compared with cars or SUVs weighing about the same. “There’s no ready explanation for this exception,” Lund says. “It probably has some thing to do with how light pickups are driven and their use patterns compared with larger and heavier pickups.
Rates differ among similar vehicles: Be-sides these broad death rate differences across vehicle groups, the rates varied within body style and size groups. In almost every size group of two-door and four-door cars, for example, the death rate for the worst vehicle was at least twice as high as the rate for the best vehicle (see tables, pp. 4-5).
Consider the Infiniti G20’s rate of 46 deaths per million registered years, which was much -lower than rates for other small four-door cars. The Chevrolet Cavalier’s rate was 162 per million, and the Pontiac Sunfire’s was 160. The upper confidence bound for the G20’s death rate is well below the lower confidence bounds for the other two cars. “This means that the lower death rate for the G20 wasn’t due to chance,” Lund says.
A more extreme example involves midsize four- wheel-drive SUVs. The Toyota 4Runner had only 12 driver deaths per million registered years during 2000-03. This compares with 134 deaths per million for the two door Ford Explorer and 119 per million for the Land Rover Discovery Series II.
Single- versus multiple-vehicle crashes: In many vehicle groups, driver death rates are split fairly evenly between single and multiple-vehicle crashes. But there are exceptions. Most driver deaths in large four-door cars and minivans occurred in crashes involving other vehicles. In contrast, in pickup trucks and SUVs of almost every size more deaths occurred in single-vehicle crashes.
In large four-wheel-drive SUVs, for example, the death rate was almost three times as high in single-vehicle crashes as it was in collisions involving two or more vehicles (14 deaths per million compared with 40).
Rollover crashes: Eleven vehicles, all pickups or SUVs, had more than 75 driver deaths per million in single-vehicle rollover crashes. This is in large part because pickup trucks and SUVs have relatively high centers of gravity compared with cars. The Ford Excursion is a very large SUV with a high rollover death rate. This is at least in part because its occupancy rate tends to be high, which raises its center of gravity even higher.
The vehicle with the very highest driver death rate in single-vehicle rollover crashes was the two-door, two-wheel-drive Chevrolet Blazer. The 251 deaths per million for this SUV compare with an average of 63 for all midsize two-wheel-drive SUVs, 34 for four-wheel- drive versions, and 28 for all vehicles in the study.
Not all midsize SUVs had high death rates in single-vehicle rollovers. The Lexus RX 300, Toyota 4Runner, Nissan Pathfinder, and Acura MDX had 6 or fewer rollover deaths per million vehicle years. Both the RX 300 and the 4Runner are equipped with electronic stability control, which has been shown to significantly reduce the risk of fatal single-vehicle crashes including rollovers (see Status Report, Jan. 3, 2005; on the web at www.iihs.org).
Not one driver death occurred in a rollover of the RX 300 or four-wheel-drive Toyota RAV4, a small SUV. This experience doesn’t mean the rates for these vehicles will be zero every year, but it does mean very low rates can be expected. “Small SUVs have had high rollover death rates in previous years, but as the RAV4 indicates this may be changing. One reason may be that the drivers are changing,” Lund points out. “It used to be that younger people, especially young men, drove small SUVs, but now many women drive them, including older women. In addition, as vehicle manufacturers redesign their small SUVs they’re addressing the rollover problem in the designs, making these vehicles more stable and less likely to roll over.”
How the death rates were computed: Rates of driver death in all crashes plus rates in multiple-vehicle, single-vehicle, and single-vehicle rollover crashes were computed for 199 passenger vehicle models (1999-2002) with at least 120,000 registered vehicle years or 20 driver deaths during the study years.
Each model’s rate represents the reported number of driver deaths divided by the model’s number of registered years. Data are from the federal government’s Fatality Analysis Reporting System and registration counts from The Polk Company.
Among the vehicles, exposure varies considerably. For example, the number of registered vehicle years for midsize two-door cars is nearly 3 million. This compares with fewer than 300,000 registered years for large two-door cars. Because of this variability, 95 percent confidence intervals were computed with upper and lower bounds indicating the precision of the computed rates for all crash types. The rates reflect primarily the influence of a vehicle’s design and patterns of use. Because driver demographics can be a major influence, the death rate for each vehicle was adjusted according to the proportion of deaths of women 25-64 years old. These drivers are involved in fewer fatal crashes per licensed driver. For most vehicles the rates were adjusted by less than 20 percent. “This is the first year we’ve adjusted the rates to account for some driver characteristics,” Lund says. “The adjustment takes away some of the differences among vehicles caused by differences in driver gender. Other demographic factors still influence the death rates, but more of the differences in the rates reflect the vehicles.”