Traffic Deaths Dropped 4%. Two Age Groups Got Worse.

NHTSA estimated 39,345 traffic deaths in 2024, down 3.8% from 40,901 in 2023.^[1] Every age cohort between 15 and 64 contributed to that decline, with the 25-to-34 bracket dropping 8% and the 55-to-64 bracket dropping 7%, five consecutive groups all heading in the same direction, all shedding deaths at rates that would make any traffic safety researcher reach for the champagne.

Two brackets did not.

Children younger than 15 saw fatalities rise from 1,023 to 1,038, while adults 65 and older climbed from 7,923 to 8,081.^[1] Combined, those two groups added 173 deaths in a year that shed 1,556 everywhere else. Plot every age bracket on the same axis and five lines slope down while two slope up, and no amount of smoothing, binning, or seasonal adjustment fixes that divergence.

Vehicle age explains why the gap exists, and the data is unambiguous about the mechanism. Occupant fatalities in passenger vehicles younger than 10 years old fell 8%, but in vehicles 10 years or older the drop was only 5%.^[1] AEB, updated ESC calibrations, and improved crash structures are doing measurable work, but only for the people sitting inside vehicles that actually have them. Children under 15 do not pick their own vehicles, and seniors 65 and older keep theirs the longest, which means both groups are disproportionately riding in the aging fleet that newer safety tech has not yet reached.

FARS model-year data confirms the mechanism: a 2004 Toyota Camry accumulated 299 deaths across the FARS decade, a 2018 Camry accumulated 148, and a 2022 Camry just 33.^[2] Same nameplate, same roads, radically different outcomes tied almost entirely to which generation of crash structure absorbs the impact. Multiply that pattern across every model in the fleet and you get a safety dividend that flows disproportionately to whoever drives the newest metal.

I will grant the counterargument: +1% for under-15 is 15 deaths on a base of 1,038, within the margin of error for the ratio-adjusted early estimates that NHTSA itself flags as preliminary, and that objection is fair as far as it goes. But +2% for 65+ is 158 additional deaths on a base of 8,081, and the 65+ population is growing while their share of newer vehicles is not growing at the same rate. That gap is widening, not closing, and the preliminary label does not change the direction.

If you have kids who ride in a car older than 2016, it predates mandatory AEB; if your parents still drive a 2012 anything, it predates updated side-impact standards. Safety tech is working, but it is not working for everyone equally, and the gap shows up exactly where you would expect it: at the two ends of the age distribution that have the least control over what they drive.