58,186 Americans Died Because They Picked the Wrong Model. Not the Wrong Class. The Wrong Model.

Here’s a fun fact that will ruin your morning commute. I pooled the death rates for every vehicle class in the FARS database (sedans, SUVs, pickups, sports cars, vans) and computed how many people would have died in each model if it performed at its class average. The gap between what actually happened and what should have happened is 58,186 corpses over ten years.

That sounds like an accusation, and in a way it is, just not aimed at anyone in particular. What this calculation does is put a body count on model-level variation: 30.4 percent of all FARS occupant deaths exist because Americans chose a specific model instead of the average model in the same class, not the wrong class but the wrong sedan within their own preferred category.

Honda’s Accord leads with 3,879 excess deaths: at a rate of 3.07 per 100 million VMT against a sedan class average of 1.39, 7,102 Accord occupants died when 3,223 would have at class-average performance.^[1] Chevrolet’s Silverado adds another 2,790 excess on top of that, and its discontinued Impala contributes 2,723 more at a rate of 5.0, still killing at three and a half times the sedan average years after GM stopped building it.

On the other side of the ledger, vehicles below their class average collectively saved 58,167 lives across the same decade.^[1] The Ram 1500 accounts for 4,309 of those saved lives with a rate nearly seven times lower than the pickup class average, and the Tesla Model 3 accounts for 2,430 at a rate almost twenty-eight times below the sedan average. Every one of these sat on a dealer lot next to something that would have killed the buyer at twice the rate.

Flatten 5,819 annual excess deaths against the most recent NHTSA estimate of 39,254 total road fatalities and you get nearly 15 percent.^[2] That share of American traffic deaths could be eliminated not by autonomous vehicles, not by AEB mandates that take eight years to phase in, but by people buying a different model at roughly the same price point. A RAV4 costs about the same as a Ford Explorer, and the Explorer has 2,451 excess deaths hanging from its rear bumper like tin cans on a wedding car bound for a funeral.

The counterargument deserves its full strength: vehicle choice correlates with driver demographics, and some of the Accord’s excess may trace to who drives Accords rather than to what happens when one hits a guardrail.^[1] This excess-death number is a ceiling, not a floor, and some fraction would persist even if every Accord became a Camry overnight. But IIHS research published in February 2025 found that safety benefits from additional vehicle weight plateau above roughly 4,000 pounds while the danger heavier vehicles pose to crash partners keeps climbing,^[3] and the rate gap between the Accord at over three per 100M VMT and the Prius at roughly half that is far too wide to be explained entirely by who sits behind the wheel.

What this changes: Next time someone tells you SUVs are safer than sedans, remember that the worst SUV in FARS (the Chevrolet Tracker at 7.83) is deadlier than the worst sedan, and that class is a marketing category while model is a survival prediction. Compare death rates within your preferred class at iihs.org/ratings, because the gap between the best and worst model in any class is wider than the gap between the best and worst class.