America’s Road Death Rate Just Broke Below 1.0. It Did the Same Thing in 2014.

NHTSA announced today that the first-quarter 2026 fatality rate dropped to 0.99 deaths per 100 million vehicle miles traveled, the lowest quarterly figure in twelve years.^[1] Traffic deaths fell 4.3% to 7,770 in the first three months of the year, and headlines will use words like "milestone" and "historic."

The Q1 2014 fatality rate was also 0.99.^[2] Same number, different decade. In the intervening years, Americans added roughly 500 billion annual vehicle miles to the road and put 50 million more registered vehicles into service, while the auto industry rolled out automatic emergency braking, lane-departure warnings, adaptive cruise control, and a fleet-wide transition from sedans to SUVs that reshaped the physics of every crash. After all of that engineering and all of those miles, the rate landed precisely where it started.

Quarters lie, and Q1 lies loudest. It is always the best quarter for road safety because winter suppresses discretionary driving, keeps motorcycles in garages, and shortens the daylight hours that encourage the long highway trips where fatigue and speed conspire against you. In Q1 2015, the rate posted at 1.03, then climbed to 1.16 by Q3. In 2020, Q1 looked encouraging right up until the pandemic emptied the roads and the remaining drivers decided speed limits were advisory. Seasonal patterns are not trends but recurring optical illusions that regulators announce as progress and journalists repeat as news every spring.

No annual rate has ever finished below 1.0 — not once in the history of the metric. In 2014, the lowest full-year figure on record landed at 1.08.^[3] Every year since has been higher. COVID pushed it to roughly 1.34 in 2020. Recovery since then has clawed the annual figure back toward 1.1, which is where it sat before the pandemic hit. We are recovering to the trend line, not breaking through it.

What actually drove the decline is less flattering than the number suggests. Between 2014 and 2026, new-vehicle sales shifted dramatically from sedans to SUVs and crossovers: sedan share dropped from about 45% to under 20%. In FARS fatal-crash data from 2014 through 2023, sedans convert crashes to deaths at a rate of 64.5%. For every 100 sedan fatal crashes, roughly 65 people die, whereas SUVs convert at 52.4% and pickups at 48.9%.^[4] Swap the fleet composition from a sedan-heavy mix to an SUV-heavy one, and the aggregate death rate drops not because any individual vehicle got safer, but because the vehicles doing the crashing are structurally more survivable for their occupants. Genuine improvement, driven by a vehicle-class substitution rather than a behavioral revolution or a regulatory triumph.

Meanwhile, the people outside those larger vehicles are paying the price. Pedestrian fatalities have risen substantially since 2014, climbing from around 4,900 annually to over 7,500 by the early 2020s, an increase that tracks almost perfectly with the growth in SUV and pickup market share.^[5] Taller, heavier vehicles protect their occupants well in a crash, but they also turn pedestrian collisions from injuries into fatalities — hood height is a death sentence, and a fleet composition shift toward bigger, taller vehicles improves the per-VMT rate for drivers while worsening the per-VMT rate for everyone who walks.

There is a version of this story where 0.99 is a harbinger of genuine, sustained improvement. AEB mandates are now standard on all new vehicles, advanced airbag systems have matured considerably since 2014, and lane-keeping systems intervene before the drowsy drift becomes a head-on. If the annual rate for 2026 finishes below 1.05, it would represent real progress toward the first sub-1.0 annual rate in NHTSA's history, and that outcome is genuinely possible.

But the pattern has always been the same: a strong Q1 triggers optimistic headlines, summer driving erases the gains, and the annual number settles somewhere between 1.08 and 1.15, give or take the severity of the holiday drinking season and whatever policy flavor-of-the-year Congress was pretending to care about that cycle. Twelve years is long enough to know what seasonality looks like, and this particular 0.99 looks exactly like the last one. A number that rewards the calendar, not the country.

What you should do: If you're buying a vehicle, the fleet-composition data is clear: all else equal, a midsize SUV or crossover with AEB, lane-keeping assist, and a five-star NCAP rating gives you the best statistical odds in a crash. Check your specific model at iihs.org/ratings and look for Top Safety Pick+ designation. If you walk or bike in your commute, treat every SUV and pickup as a vehicle whose driver's sight lines start above your head. A fleet that makes the per-VMT rate drop is the same fleet that makes the per-pedestrian rate climb.