← All articles

🚗 Transport

At Britain’s Per-Mile Death Rate, About 21,700 Fewer People Would Have Died on US Roads in 2025

NHTSA called America's 2025 road death rate the second-lowest on record, but per mile it was still 2.45 times Britain's, a gap worth about 21,700 lives by our math.

Illustration: an aerial view at dusk of a two-lane rural highway curving across dark farmland toward a small lit town, with white headlight and red taillight trails, short side roads to lit farmyards, and reflector posts along the shoulder.

If an American mile had been as deadly as a British one in 2025, about 21,700 fewer people would have died on US roads. That is nearly 6 in 10 of the 36,640 road deaths the National Highway Traffic Safety Administration estimates for the year, or about 59 people a day, and it comes in a week of good news from the agency.

On November 10, NHTSA reported a first-half 2026 death rate of 1.02 per 100 million vehicle miles, which it called a near-record low, and in April its 2025 estimate had called that year’s rate of 1.10 the second-lowest in recorded history. Britain’s Department for Transport counted 1,538 road deaths in 2025 in its final results, and its traffic estimates put British driving at 342.6 billion vehicle miles, which works out to 0.45 deaths per 100 million miles: one death every 223 million miles, against America’s one every 91 million.

Driving More Is the Smaller Factor

Per million residents, the United States lost 107 people in 2025, using the Census Bureau’s Vintage 2025 estimate of 341.8 million; the United Kingdom lost 23.3, DfT reports. That 4.6-fold gap is two factors multiplied together. These figures imply Americans log about 1.9 times as many vehicle miles per resident, and each of those miles is 2.45 times as deadly (1.9 × 2.45 ≈ 4.6). Had Americans driven only as far as Britons at their own per-mile risk, deaths would have been roughly 19,500 instead of 36,640, a big drop that still leaves the toll per person at about 2.45 times Britain’s. On a multiplicative scale, distance explains a bit under half of the gap and per-mile risk a bit over half.

The Gap Holds on City Streets and Country Roads

DfT’s road-type tables give deaths and traffic for each kind of road; the Insurance Institute for Highway Safety publishes the same split for the United States, from federal crash records for 2024.

Road deaths per 100 million vehicle miles, United States and Great Britain
Road typeUnited StatesGreat BritainUS multiple
All roads (both 2025)1.100.452.45×
Urban, excluding GB motorways (US 2024, GB 2025)1.010.442.3×
Rural, excluding GB motorways (US 2024, GB 2025)1.560.602.6×

British road-type rates are our calculation from DfT’s deaths (527 urban, 923 rural) and traffic (119.2 and 155.1 billion vehicle miles); that table sums to 346.2 billion miles, slightly more than the 342.6 behind the all-roads row. Two quirks pull in opposite directions. First, American rates are a year older, and if they fell as much as the national rate did in 2025, the multiples would be about 2.1 and 2.4. But DfT counts motorways separately, while the American urban and rural figures include interstates and freeways, which should pull the US rates down and makes these multiples conservative. An American city mile is more than twice as deadly as a British city mile. So is a country mile.

The Calculation

NHTSA’s 2025 figures imply 3.33 trillion vehicle miles (36,640 deaths divided by 1.10 per 100 million), and at Britain’s rate of 0.449 those miles would have produced about 14,950 deaths, a gap of roughly 21,700. Rounding in the published rate moves that between 21,600 and 21,800, DfT’s road-type mileage would make it about 21,800, and revisions to NHTSA’s early estimate could move it further. Britain is a demanding benchmark, though, so we also ran gentler ones. At 3, 4 and 5 deaths per billion vehicle-kilometres (Britain converts to about 2.8), the 2025 gap would be about 20,600, 15,200 and 9,800. None of them is small.

How Long to Catch Up

In the first half of 2026, deaths fell from 17,175 to 16,750 while driving rose 0.8 percent, so the rate fell about 3.25 percent (NHTSA’s rounded 1.05 and 1.02 give 2.9). At that pace America would reach Britain’s 2025 rate in about 27 years, or 30 at 2.9 percent; at 2025’s 7.6 percent drop, about 11. That is a generation. Both are straight-line extrapolations, neither assumes Britain improves, and the death count is barely moving: NHTSA says deaths in the second quarter of 2026 fell just 0.2 percent.

What You Can Do

Drivers: the 2.45 is a national average, not your personal odds, but the basics still matter most where the US rate is highest, on rural roads (1.56 per 100 million miles in 2024): buckle every seat, hold the limit on two-lane highways and don’t drive impaired.

City and county officials: ask your state transportation department for local deaths per vehicle mile, averaged over three to five years because single-year counts swing. Put them next to Britain’s 0.44 urban figure, which excludes motorways. Definitions differ, so treat it as rough. Still, a rate near 1.0 is the American norm, not a good result.

Reporters: when an agency calls a rate near-record, ask what it is near to. A domestic record measures progress, not the distance left.

Limitations

US figures for 2025 and 2026 are NHTSA early estimates and will be revised, and both countries count a death within 30 days of a crash, under NHTSA’s FARS criteria and DfT’s definitions. Urban and rural are drawn differently: DfT uses towns of 10,000 people or more (3,000 in Scotland), while American boundaries are set by state highway departments and approved by the Federal Highway Administration, according to IIHS, and the per-person figure is for the whole United Kingdom and the per-mile figures for Great Britain. The 21,700 is an arithmetic benchmark against one of the safest road systems measured, not an estimate of lives any particular policy would save, and nothing here identifies why American miles are deadlier, whether speed, vehicle size, road design or enforcement.

The Strongest Case Against This Analysis

Here is the best case against the comparison. Britain is a cherry-picked benchmark: DfT’s main findings rank it fourth-lowest by deaths per head among countries with 2025 data. By distance it does about as well: the International Transport Forum’s 2025 road safety report (linked here through a third-party copy) charts 22 countries by deaths per billion vehicle-kilometres and names only Norway (1.9), Sweden (2.6) and Denmark (2.9) below 3, so Britain’s 2024 figure there was somewhat higher, but its 2025 rate converts to about 2.8, which puts it at the safe end either way. American roads are also different in kind, with long, fast rural two-lane highways that Britain largely lacks. Fair enough. That is why the 21,700 should be read as the gap to a leader, not to an average. But the same report puts the United States tied for the highest rate, 7.8, with Czechia, in the latest year each reported, and even against a benchmark nearly twice as deadly as Britain’s, the gap stays near 10,000 deaths a year. And the road-mix objection fails on its own terms, because the gap survives within urban and within rural roads.

The Bottom Line

NHTSA is right that American roads got safer in 2025 and again in early 2026, but measured against Britain the second-lowest rate in recorded US history still meant roughly 21,700 more deaths than British per-mile risk would have produced, about 59 a day, on city streets and country highways alike. At this year’s pace, catching Britain takes 27 to 30 years. The number to watch is deaths per mile against peer countries, not the domestic record.

Related