Somewhere in the foothills of central Virginia, there’s a nonprofit that has spent more than sixty years perfecting one very specific skill: figuring out exactly how much your car is going to hurt somebody. This week it published its latest report, and a lot of the coverage treated it like a bulletin from a federal safety regulator. It isn’t one. It never has been.
That distinction matters more than any number in the new driver death rate study from the Insurance Institute for Highway Safety, and there are some brutal numbers in it. The Kia Rio posted the worst driver death rate of any 2023 model year vehicle, at 170 deaths per million registered vehicle years. A Ram 3500 crew cab long bed four-wheel-drive posted the worst other-driver death rate, at 204. Auto Wire already broke down what those numbers mean for the physics of weight and crash compatibility, and separately what they mean for the six vehicles that reported zero driver deaths at all. Read both if you haven’t. This isn’t a rehash of that math. This is about who’s been keeping score, and why they started keeping it in the first place.
IIHS Is Not The Government, And That’s Not A Technicality
People conflate the Insurance Institute for Highway Safety with the National Highway Traffic Safety Administration constantly, and it’s an understandable mistake. Both crash cars into barriers. Both hand out star ratings that show up on window stickers. Only one of them is a federal agency. IIHS was founded in 1959 by three insurance trade associations that together represented roughly 80 percent of the American auto insurance market at the time. Its original job wasn’t independent research. It existed to support highway safety campaigns other people were already running. The Institute didn’t become the research organization we recognize today until a decade later, when William Haddon Jr., the nation’s first federal highway safety chief, took over as its president in 1969 and rebuilt it around actual science.
That’s a remarkable pedigree for an organization whose ratings get treated like federal law. They aren’t. They were built, and are still funded, by the industry that writes the check after the crash.
The Sister Organization Whose Whole Job Is Insurance Math
IIHS doesn’t operate alone. Its companion organization, the Highway Loss Data Institute, shares office space and a mission statement. HLDI’s job, in IIHS’s own description, is studying insurance losses and publishing “loss results by vehicle make and model”. That isn’t a research mission with a side interest in insurance. That’s an actuarial mission, stated plainly, on the same page that introduces the crash tests everyone assumes are pure science.
None of that makes the death rate numbers fake. The methodology is public, the sample sizes are disclosed, and IIHS is unusually candid about its own blind spots. Its release openly admits the rates aren’t adjusted for how many miles people drive or how fast, and that cheaper cars likely attract younger, higher mileage drivers regardless of how the sheet metal performs. That candor deserves credit. But candor about the numbers isn’t the same as neutrality about why the numbers get collected. An organization built by insurers to quantify loss is going to be very good at quantifying loss. That’s not a scandal. That’s a design specification.
The Blind Spot Nobody’s Star Rating Admits To
IIHS’s crash tests and NHTSA’s federal 5-Star ratings share a limitation that almost never makes it into headline coverage. Frontal crash results, NHTSA explains on its own ratings page, can be compared only between vehicles from the same weight class. A five-star Kia Rio and a five-star Chevrolet Tahoe are not making the same promise. The barrier test measures how a vehicle absorbs its own kinetic energy, and kinetic energy is a function of mass. It cannot tell you whether a 2,600-pound car is safer than a 5,700-pound SUV, because it was never built to answer that question. Death rates can compare across weight classes. Star ratings can’t. Auto Wire has already shown how that same weight math draws a hard, uncomfortable line for teen drivers shopping used cars. Most shoppers have no idea they’re reading two completely different instruments.
Which brings us to the Ram 3500. Search for it in IIHS’s own consumer ratings tool and nothing comes up. Not a poor score. No score at all. The truck responsible for the highest other-driver death rate in the entire fleet, the one statistically most dangerous to whoever it hits, has never been run through the barrier test everyone assumes generates this kind of authority. Heavy-duty pickups routinely exceed the weight window the standard passenger-vehicle protocol was built around. The most consequential vehicle on this list is also the one the rating system was never designed to grade.
Why This Outlasts The News Cycle
This year’s release marks only the second time IIHS has calculated an other-driver death rate at all, the metric that flips the entire safest car conversation by asking what a vehicle does to people who never bought it. That metric is new, it’s getting more sophisticated with each cycle, and it lives inside an organization whose sister institute exists specifically to turn loss data into numbers insurers can price. Top Safety Pick badges, the same ones Audi, BMW, Mazda, and Subaru fought hard to win this summer, already move the needle on discounts at some insurers. There’s no reason to assume the other-driver metric stays a research curiosity once it’s been tracked long enough to be actuarially useful. That’s not a conspiracy theory. It’s just what this particular institution has always eventually done with a good number.
A crash test looks like pure science because it happens inside a laboratory. It’s worth remembering who built that laboratory, and who’s still paying the electric bill six decades later.