America's Traffic Death Rate Just Hit a Record Low. It's the Same Number From 2014.

NHTSA's early estimate for 2025 pegs the U.S. traffic fatality rate at 1.10 deaths per 100 million vehicle miles traveled.^[1] Fifteen consecutive quarterly declines, down 6.7% year over year, 39 states plus D.C. and Puerto Rico trending in the right direction. The press release practically vibrated with satisfaction.

In 2014, the rate was 1.08.^[2]

That is the whole story, not a preamble. A decade of ESC mandates reaching full fleet saturation, automatic emergency braking migrating from luxury option to federal rulemaking agenda, $665 million in annual safety grants distributed to all 50 states, and an entire generation of highway engineering research bought the country two hundredths of a point.^[1] A rounding error with a press conference.

Look at a chart, though, and you will see something more flattering: a crisis followed by a triumphant recovery. Rate held at 1.10 in 2019, then detonated to 1.34 in 2020 and 1.37 in 2021 as COVID hollowed out highways and replaced commuters with reckless, often impaired drivers.^[2] Twenty-one percent rate increase in a single year. Watching the subsequent four-year decline from 1.37 back to 1.10 looks genuinely impressive on a graph, provided nobody checks what the rate was before the spike happened. Celebrating "lowest since 2019" required collective amnesia about the fact that 2019 was already worse than 2014.

I calculated the price of that amnesia. If the pre-COVID rate had held through 2020 to 2024, roughly 30,000 fewer Americans would have died on roads.^[2][3] Nearly 7,000 excess deaths in 2020 alone when VMT plummeted but recklessness spiked, peaking above 8,000 in 2021, then tapering to roughly 3,000 in 2024 as the rate slowly bled back toward its pre-pandemic baseline. More American dead across those five years than the Korean War's first twelve months produced. What NHTSA is celebrating is regression to a mean the country should never have abandoned.

The Strongest Case Against This Reading

Consider what happened to American driving since 2014. Smartphones turned distracted driving into a mass-participation epidemic. Average vehicle curb weight climbed roughly 300 pounds as consumer preference lurched toward trucks and large SUVs, which surged from half of new sales to nearly 80%, putting heavier, taller front ends at pedestrian head height.^[4] Rideshare platforms injected billions of additional VMT from drivers navigating unfamiliar routes while watching phone-mounted GPS screens. Against that headwind matrix, holding 1.10 is not failure. It is a massive, quiet technological intervention preventing a demographic slaughter that every behavioral trend predicted.^[5]

That argument is genuinely strong, and it might be correct. But accepting it means the ceiling of American traffic safety ambition is "didn't get worse." Over the same decade, EU member states cut their per-capita fatality rate by roughly 17%, from about 51 deaths per million inhabitants down to 45.^[6] Sweden's Vision Zero corridors halved deaths in treated areas. Those countries faced identical smartphone proliferation, identical vehicle-weight inflation, identical rideshare disruption. They found real improvement while America found a treadmill and called the sweat a personal best.

Limitations

VMT figures for 2020 through 2023 are derived from published death counts divided by published rates, not direct federal measurement; margins of ±5% apply, especially for 2020 when FHWA recorded a sharp drop to roughly 2.83 trillion miles.^[3] The excess death calculation assumes the 2019 rate as the counterfactual, which is generous to the official narrative: using 2014's 1.08 would widen the gap further. Some behavioral deterioration during the pandemic, particularly increased substance abuse and higher risk tolerance among younger drivers, may prove to be a permanent shift rather than an anomaly that will fully revert. Rate per 100M VMT does not capture per-person or per-trip risk, which matters more for individual decision-making than aggregate national statistics.

What You Can Do About It

Your individual risk on American roads has not materially changed in a decade if you drive a similar vehicle on similar roads. To beat the national average: buy or lease a post-2018 vehicle with AEB, which reduces rear-end crash involvement by roughly 50% according to IIHS research.^[5] Avoid driving between 2 and 4 AM on weekends, the deadliest window in FARS data across every year studied. Run your VIN at nhtsa.gov/recalls today, not tomorrow, because 53 million vehicles in this country currently carry unfixed recall defects.^[1] And the single most effective safety upgrade available remains what it has been since FARS started counting bodies: do not combine alcohol with driving, a factor in more than 30% of all traffic fatalities every single year.^[2]

Source: NHTSA FARS 2014–2023, NHTSA 2025 early estimates, FHWA VMT data. Excess death figures are author calculations using derived VMT estimates with ±5% uncertainty. See methodology for caveats.