← The Crash Report
Existential Dread

NHTSA Called 39,345 Deaths ‘Encouraging.’ Norway Lost 89 People.

NHTSA just dropped its 2024 full-year estimate, and America’s federal highway safety agency used the word “encouraging” to describe 39,345 dead people.[1] That figure is the lowest since 2020, which tells you more about what happened in 2020 through 2023 than anything encouraging about 2024.

24,045
excess American road deaths per year above the EU average

Across the Atlantic, the EU published its own 2024 numbers the same week.[2] The entire European Union lost 19,940 people on its roads. That continent has 450 million residents, and America, with 340 million, managed to kill twice as many of them in cars, trucks, and crossovers that weigh a combined several hundred million tons more than their European equivalents.

Norway, a country of 5.5 million people who drive on mountain roads in the dark for five months of the year and whose national speed limit is 80 km/h on highways, lost 89 people. Eighty-nine. That is a rate of 16 per million population. Sweden, the country that invented Vision Zero and actually followed through, lost 213 at a rate of 20 per million. Denmark managed 24 per million in a nation where it rains sideways through most of October, and the Dutch, who share roads with more bicycles than humans, hit 32 per million.

America’s rate sits at roughly 116 per million. If you need that translated into something concrete: for every person who dies on a Norwegian road, seven die on an American one, adjusted for population, and the Norwegian had worse weather, narrower lanes, and a government that treats 89 deaths as a national failure rather than an “encouraging” quarterly trend.[3]

22.5%
increase in U.S. death rate 2013–2022, while 27 peer nations dropped 19.4%

The CDC ran this comparison for 2013 through 2022, and the finding would be comical if it weren’t measured in bodies: the United States increased its road death rate by 22.5 percent over the decade while the median high-income peer nation decreased by 19.4 percent.[4] The two curves are now moving in opposite directions at a combined divergence of 42 percentage points per decade, which is the kind of chart you show to a congressional committee if you want them to stare at their shoes for twenty seconds before changing the subject. Pedestrian deaths are worse: the US rate climbed 50 percent while peers dropped 25 percent, producing the highest pedestrian fatality rate among 15-to-64-year-olds of any country in the study, a distinction that no amount of reflective running gear or pedestrian-detection cameras has managed to reverse.

What would it take to match the European average? Nothing exotic. If the US achieved the EU’s 45-per-million rate, we would expect roughly 15,300 deaths per year instead of 39,345, a reduction of 24,045, which works out to a 9/11-scale event every 45 days that simply does not happen in countries where roundabouts outnumber stroads and the default speed limit in residential areas is 30 km/h, not whatever the traffic engineer decided was the 85th percentile of what drivers were already doing. Matching Norway would save 33,900, but let’s not get greedy.

NHTSA Chief Counsel Peter Simshauser acknowledged, almost parenthetically, that “America’s traffic fatality rate remains high relative to many peer nations.” The diplomatic restraint of “high” deserves a medal. A rate 2.6 times the continental European average and 7.3 times Norway’s is not high. It is structurally aberrant, the predictable output of a system that treats road design as a traffic throughput problem, vehicle size as a consumer preference, and 40,000 annual deaths as a stubborn but essentially acceptable cost of doing business across 4.2 million miles of asphalt.[1]

Strongest counterargument, at full strength: the US is not Norway, and pretending otherwise is lazy. Americans drive nearly twice the annual miles per capita as Europeans, at higher average speeds, across a landmass where the distance between Topeka and the nearest trauma center can exceed the entire width of Belgium. Rural fatality rates everywhere are higher than urban ones, and the US has more rural VMT than any peer nation. Normalizing for vehicle miles traveled narrows the gap somewhat. Fair enough. But it does not close it, and it certainly does not explain why the gap is widening while peers improve and the US regresses. Distance is a constant; policy is a variable.

Limitations: EU countries use a 30-day post-crash mortality window; FARS methodology is similar but not identical. Small nations like Norway are subject to year-to-year volatility. The 2024 US figure is a preliminary estimate; final data arrives 12 to 18 months later. Per-VMT comparisons are imperfect because not all countries report VMT the same way.

What you can do: the gap is a policy choice, not geography. If you drive, check whether your state has adopted a Complete Streets policy (most haven’t). Support local speed limit reductions in residential areas. The European playbook is not a secret: narrower lanes, lower default speeds, roundabouts over signals, separated bike infrastructure, and vehicle design standards that penalize front-end height. None of this requires inventing new technology. It requires deciding that 39,345 is not, in fact, encouraging.

Sources & References

  1. NHTSA, “NHTSA Releases 2023 Traffic Deaths, 2024 Estimates,” June 2026. nhtsa.gov
  2. European Commission, “Road Safety Statistics for 2024: Progress continues amid persistent challenges.” transport.ec.europa.eu
  3. NHTSA, “Traffic Crash Deaths: Early Estimates Jan–Sept 2024.” nhtsa.gov
  4. CDC MMWR, “Pedestrian and Overall Road Traffic Crash Deaths — United States and 27 Other High-Income Countries, 2013–2022,” Vol. 73, No. 49. cdc.gov

Source: NHTSA 2024 preliminary estimate, EU CARE database 2024 provisional data, CDC MMWR 2013–2022 analysis. US 2024 figure is a preliminary estimate; final FARS data available 12–18 months later. EU data uses 30-day mortality window; slight methodological differences exist. Small-nation rates are volatile year to year. See methodology for caveats.