America Just Spent Six Years and 24,000 Lives Getting Back to Square One
NHTSA's 2025 fatality estimate arrived in April carrying fifteen consecutive quarterly declines and a congratulatory tone: 36,640 deaths, a rate of 1.10 per 100 million vehicle miles traveled. Same rate as 2019, the second-lowest in recorded history behind only 2014. The agency described this as “back to prepandemic levels,” a phrase that frames a six-year roundtrip costing approximately 24,000 extra American lives as a destination rather than a damning indictment of everything that went wrong between March 2020 and the day someone in Washington decided the chart looked acceptable again.[1]
Between 2020 and 2025, American roads killed 241,877 people. At the 2019 baseline of 36,355 deaths per year, the six-year expected total is roughly 218,130, which leaves a surplus of 23,747 human beings who died inside the behavioral crisis that COVID unleashed and the country spent half a decade failing to contain: speeding that never fully abated even after traffic returned, impaired driving that spiked in 2020 and plateaued through 2022, seatbelt noncompliance that settled into habit across dozens of states with secondary enforcement laws that amount to polite suggestions.[2]
Year by year the toll was 2,652 excess deaths in 2020, then 6,875 in 2021 at the worst of it, 6,366 in 2022, followed by a grinding four-year retreat through 4,670, 2,899, and a final 285 before the rate limped home to 1.10. That is six full calendar years and 23,747 obituaries to arrive back at the same number on the same chart, and the people running the victory lap would like you to focus on the arrival rather than the cost of the trip.
No peer country experienced anything comparable. A CDC study of 28 high-income nations found that between 2013 and 2022, the United States increased its overall traffic death rate by 22.5% while every other country in the sample cut theirs by a median of 19.4%, with pedestrian fatalities diverging even more violently: up 50% domestically, down 24.7% abroad.[3] Whatever happened on American roads was not a pandemic side effect shared by developed nations that faced the same virus, the same lockdowns, the same economic disruptions; it was a failure of road design, speed enforcement, and impairment tolerance specific to a country that still treats 36,000 annual traffic deaths as an acceptable operating cost for its transportation network.
Vehicle age explains the mechanism, because FARS records from 2014 to 2023 show that 62.3% of all fatalities involve cars and trucks at least a decade old, with the 10-to-14-year cohort alone representing 27.7% of total deaths while vehicles under five years old contribute just 5.2%.[4] COVID suppressed new-car purchases for years, trapping families inside aging crash structures built before automatic emergency braking became standard equipment, and every additional season those vehicles stayed on the road compounded the fatality burden at a rate that national statistics eventually absorbed but individual households never will. Fleet turnover ended the crisis, not enforcement and not public education campaigns, because the actual fix was newer vehicles with better physics and better software reaching the fleet faster than old behavior could kill people in old cars.
A counterargument deserves full strength: the 15-quarter recovery is historically unprecedented, genuine evidence that America's road safety system can self-correct after a massive behavioral shock without requiring new legislation or a federal safety mandate, and dismissing that resilience would be intellectually dishonest. But “self-corrected after 23,747 excess deaths” is a strange formulation to celebrate when the endpoint is the same number we started with, and the per-mile fatality rate still runs roughly double that of the average high-income nation in the CDC comparison set.
What you should do with this: if you drive a 2005-to-2014 model year vehicle, you sit in the peak FARS fatality bracket, and no declining national trend line changes your individual risk profile. A 2019-or-newer vehicle equipped with AEB, ESC, and current crash architecture runs a per-mile fatality rate roughly three to five times lower than its predecessor from the mid-2000s. Run your VIN at nhtsa.gov/recalls today, not because the national numbers are alarming but because the national numbers have absolutely nothing to say about whether your particular vehicle, at your particular mileage, on your particular commute, ends up in next year's 36,640.
Sources & References
- NHTSA, Early Estimate of Motor Vehicle Traffic Fatalities and Fatality Rate in 2025 (DOT HS 813 800), April 2026. 36,640 deaths, 1.10/100M VMT, 15 consecutive quarterly declines. crashstats.nhtsa.dot.gov
- NHTSA, Fatality Analysis Reporting System (FARS), annual totals 2014–2024. Yearly death counts and VMT rates used for excess death calculation: 2019 baseline (36,355), 2020 (39,007), 2021 (43,230), 2022 (42,721), 2023 (41,025), 2024 (39,254). nhtsa.gov
- CDC MMWR, “Pedestrian and Overall Road Traffic Crash Deaths — United States and 27 Other High-Income Countries, 2013–2022,” March 2025. U.S. traffic death rate +22.5%, 27-nation median −19.4%. Pedestrian deaths: U.S. +50%, others −24.7%. cdc.gov/mmwr
- NHTSA FARS 2014–2023, model-level analysis. 62.3% of fatalities in vehicles 10+ years old; 10–14 year cohort = 27.7% of deaths; vehicles <5 years = 5.2%. nhtsa.gov