Some Cars Kill You Before Their First Oil Change

A 2021 Dodge Challenger rolls off the line in Brampton, Ontario. Eight airbags. Blind-spot monitoring. Forward collision warning. Lane departure assist. Every acronym the safety brochure can fit on a page. Average time between sale and fatal crash involvement in FARS: 1.6 years.^[1]

We’ve been telling ourselves a comforting story about traffic deaths. Old cars kill people. Rust-bucket Cavaliers with the crumple zone engineering of a filing cabinet. Twenty-year-old Rangers with no electronic stability control. The average Chevy GMT-400 in a fatal crash is 23.7 years old.^[1] Buy something new. The technology will save you.

The technology will not save you from physics, theft, or yourself.

We cross-tabulated every fatal crash in FARS (2014–2023) by the gap between model year and crash year. Call it “death age” — how old a vehicle typically is when it kills someone.^[1] Most vehicles peak between 12 and 18 years. A handful peak under five. The Kia Forte: 1.6 years. The Ram 1500: 1.2. These machines are barely broken in.

Three failure modes, one paradox

The Muscle Cars. The Challenger and Charger account for 38.4% and 26.9% of their total FARS deaths from 2019–2023 model years alone.^[1] These vehicles came with between 303 and 807 horsepower. No amount of stability control can overcome a Hellcat driver doing 140 on a suburban boulevard. Dodge marketed these cars as weapons. Buyers obliged.

The Stolen Ones. The 2021 Kia Forte — a $17,000 economy sedan — recorded 76 FARS deaths in just two to three years on the road.^[1] The Kia Soul: 26.6% of deaths from near-new model years. A USB-cable exploit that went viral on TikTok turned 2011–2022 Hyundais and Kias into the most stolen vehicles in America — NHTSA reported theft rates 8 to 11 times higher than comparable models.^[2] The root cause: Hyundai and Kia chose to omit engine immobilizers — a $5 part standard on virtually every other manufacturer’s vehicles — from millions of cars. Teenagers with USB cables were the symptom. A corporate cost-cutting decision was the disease.

The Volume Machines. The 2019 Ram 1500 recorded 156 FARS deaths — more than any other single model year of any vehicle.^[1] Some of this is pure sales volume. Ram sells 500,000+ trucks a year. But the death-age skew is steeper than volume alone explains. The brand split from Dodge in 2011 means FARS captures the Ram 1500 as a distinct nameplate only from that year forward, concentrating deaths in newer model years.

What this actually tells us

The “death age” metric reveals something the per-model fatality rate obscures: old killers and young killers fail for entirely different reasons. Old vehicles die from engineering deficits — no ESC, weak structures, aging tires. Young vehicles die from behavioral deficits — too much power, trivially easy to steal, or so ubiquitous that even small per-mile risk generates enormous absolute body counts.

A 2021 Kia Forte (estimated rate: ~2.5 deaths per 100M VMT) and a 2004 Chevy Cavalier (~2.7 per 100M VMT) kill at roughly the same rate.^[1] One has every safety feature on the market. The other has the structural integrity of a beer can. Same outcome. Completely different pathways.

The counterargument (and why it doesn’t fully hold)

There’s a real methodological caveat here. FARS spans 2014–2023. A 2019 model year vehicle has at most five years of exposure in the dataset, while a 2005 vehicle has a full decade. We are comparing asymmetric observation windows, which naturally inflates the death concentration in newer model years. The Ram 1500 brand split further compresses deaths into post-2011 model years that were previously coded as “Dodge Ram.”

This is a legitimate concern. But it doesn’t explain why the Kia Forte MY 2021 accumulated 76 deaths in two to three years of road time while the Honda Civic MY 2021 — selling in comparable volumes — did not. It doesn’t explain why the Challenger’s death age is measured in months. The observation window compresses the timeline. The behavioral signal is still real.

Safety tech saves lives. We have the ESC data to prove it.^[3] But “buy a new car and you’ll be safe” is a story we tell ourselves. The FARS data says the truth is uglier: some new cars come with their own category of lethality. The airbags deploy. The crumple zones crumple. The driver was doing 140, or was fifteen years old with a USB cable, or was one of 500,000 identical trucks on the road.

Modern safety engineering is a seatbelt in a hurricane. It works — right up until human behavior overwhelms the spec sheet. And the hurricane keeps blowing.