Two vehicles on the IIHS’s new fatality tables are the same SUV. The Volkswagen Atlas with front-wheel drive ranks among the twenty models least likely to kill its own driver, at five deaths per million registered vehicle years. It also ranks fourteenth-worst in the country at killing the other driver, at 94. Add the optional driven rear axle and that second number collapses to 13, good enough for fifth-best in the fleet.
Same body. Same crumple zones. Same airbags. Wildly different body count.
That gap is the most useful thing in the new numbers the Insurance Institute for Highway Safety published on July 30, and it’s the thing most likely to get lost behind the headline stat. Yes, the Kia Rio posted the worst driver death rate of any 2023-model-year vehicle at 170 per million registered vehicle years. Yes, a Ram 3500 crew cab long bed 4×4 is statistically the most dangerous thing you can meet on a two-lane road, at 204 other-driver deaths. But the Atlas split tells you what these tables actually measure, which is not crashworthiness. It’s crashworthiness tangled up with who bought the thing, how much they paid, where they live, and how they drive it.
What the numbers are counting
IIHS has been running this calculation roughly every three years since 1989, and the method is unglamorous in a good way. Deaths come from the federal Fatality Analysis Reporting System. Registrations come from Mobility Global. Divide one by the other, adjust for driver age and sex, and you get deaths per million registered vehicle years — one vehicle registered for one year.
Related Articles
- California’s 100-MPH Crackdown Isn’t a New Law. It’s an Old One Wearing a Radar Gun
- Alpine’s US Launch Plan Has Quietly Shrunk — Here’s What Renault’s Own Documents Show
Crucially, the rates are not adjusted for miles driven, road type, speed, or driver income. A model must clear 100,000 registered vehicle years between 2021 and 2024 or log at least 20 deaths to appear at all, which quietly excludes plenty of low-volume vehicles that would otherwise embarrass their manufacturers. IIHS treats two-wheel-drive and four-wheel-drive variants as separate models precisely because the customer bases diverge — the 4×4 costs more, and it tends to live somewhere with weather.
That’s the Atlas answer, and the same logic explains why a Ford F-150 SuperCrew 4×2 lands on the lowest driver-death list at seven while simultaneously ranking twelfth-worst for other-driver deaths at 99. Two-wheel-drive full-size trucks skew toward fleet, commercial, and high-mileage southern use. They rack up exposure the denominator can’t see.
The physics nobody legislates around
The other-driver list is a weight list wearing a badge. Ten of the twenty worst are large or very large pickups. As a class, very large pickups killed 110 other drivers per million registered vehicle years against a fleet average of 53 — and only 24 of their own.
This isn’t news to IIHS, which quantified the tradeoff last year. Below the roughly 4,000-pound fleet average, every extra 500 pounds of curb weight buys real protection: 17 fewer driver deaths per million, at a cost of one additional death to crash partners. Above that average, the trade inverts. For heavy pickups, another 500 pounds saves one of your own and costs seven of somebody else’s. The airbag and structural gains of the last two decades have eaten most of the occupant benefit of mass; what’s left is the externality.
Structural alignment already did the heavy lifting here. Beginning in 2009, under a voluntary agreement IIHS helped broker, automakers dropped the primary energy-absorbing structures on trucks and SUVs into the same vertical band as cars’ bumper beams and frame rails, so a truck front end engages a sedan’s crumple zone instead of skating over it and into the A-pillars. Car occupants used to be 90% more likely to die against a 5,000-pound SUV than against another car; now it’s 20%. Pickups went from 2.5 times as lethal to their crash partners down to a little under twice. The remaining gap is mass and geometry, and a hood leading edge above 40 inches is 45% more likely to turn a pedestrian strike fatal than one below 30.
If you own a heavy-duty crew cab, the practical read is boring but real: your liability exposure is not the same as your neighbor’s in a CR-V, and state-minimum bodily injury limits are a bad match for a 204-per-million aggressivity profile. That’s a conversation for your agent, not your dealer.
The muscle car column, one last time
The Dodge Charger HEMI 2WD is one of only four models on both worst lists — 127 for its own driver, 161 for everyone else — alongside the standard Charger 2WD, the Camaro, and, improbably, the Chevrolet Malibu. The Challenger 2WD sits at 125, the Chrysler 300 2WD at 132, and the Corvette at 82.
IIHS President David Harkey’s read is blunt: “Powerful engines and marketing that attracts aggressive drivers is the issue for sports cars and muscle cars. They’re dangerous mainly because of how they’re driven.”
He’s not guessing. IIHS made the comparison explicit in its previous round: relative to weight, the 275–305 horsepower of an entry-level pony car isn’t far off the 241–252 horsepower of an Audi A4 or Mercedes E-Class. The luxury sedans land among the best performers anyway. The variable isn’t the engine. It’s the hood scoop and everything it implies about the person under it.
Related Articles
- BMW Spider-Man Dash Ad: How It Got Into Your Car And Why Nobody Can Stop It
- Mercedes’ Next EVs Are Being Built From Scrapped Cars. The Real Reason Has Nothing To Do With Saving The Planet
There’s a wrinkle here the data can’t reflect. Brampton built its last Last Call cars in the 2023 model year, and the sixth-generation Camaro retired after 2024. Every V8 sedan on this list is a used-car proposition now, which means these cars are moving down the ownership ladder toward exactly the buyers most likely to reproduce the behavior that generated the numbers. Expect the next study to look worse for them, not better, even as the nameplates disappear from showrooms.
The trend line, finally bending
Zoom out and there’s something almost encouraging. The fleet-wide driver death rate ran 28 for 2011 models, 30 for 2014, 36 for 2017, and 38 for 2020. This round: 37. It’s the first decline after a decade of climbing, and it arrives before the federal automatic emergency braking standard even bites — FMVSS No. 127 doesn’t require compliance until September 1, 2029, so essentially none of the 2023 fleet was built to it.
Very large pickups also improved as a class, from 121 other-driver deaths to 110. The Ram 3500 crew cab long bed 4×4 did not: it topped this list last time at 189 and topped it again at 204.
Minicars got worse too, from 153 to 158. And that’s the uncomfortable arithmetic underneath all of it. The cheap small car is the most dangerous seat in America, the cheap small car is what you can afford if you’re new, young, or working two jobs, and the industry solved the problem by discontinuing the cheap small car rather than fixing it. The full model-by-model table is worth twenty minutes before your next purchase — just read both columns, not one.