Road fatalities per 1 billion vehicle-km (factors out the distance driven): USA is 7.6; Sweden is 3.7. I'm in the UK, we're 2.6.
Road fatalities per 100,000 inhabitants per year (a decent measure of "how likely I am to know someone who has died on the road"): USA is 11.6; Sweden is 3; UK is 2.7.
Urban form and the external cost of transportation are inextricably linked. You can't say that of course it's safer in Sweden because they drive less, because it's safer in Sweden _because_ they drive less! That's the point!
The USA has chosen to require almost everyone to drive long distances with low density zoning, poor public transit, and extremely high parking mandates on private landowners.
If another country has lower fatalities mostly because the government doesn't make them drive as much, that is a result of public policy, just as the high rate of driving in the USA is.
Also, the USA has extensive investments in road safety, probably more than any other nation. Modern freeway safety design standards to enable high speed driving without as many fatalities originates in the USA and close to US$100 billion each year is spent on high speed roads and improvements.
Also, the USA invests extensively in the world's best trauma surgery corps. The level of fatalities on the roads would probably be double what it is with merely typical first world medicine. Emergency trauma surgery to fight gunshot and car crash deaths is one of the few fields of medical care where the USA far exceeds world standards.
It'd be like measuring ad success rate by ad size instead of clicks.
We have close to taxation of Ricardian land rents, but probably not enough. Even if it's not absolutely true, Georgist land rents taxation is very interesting nonetheless.
This is all good. The swedish government has been spending ridiculous amounts of money building away death in traffic.
But, I suspect the americans probably drive a lot more than the average swede. I think if you change the death toll from per-person to per-driver the numbers will change and if you change it to per-driven-mile I'm not sure there will be a difference any more. Anyone know if there are numbers on this? Can't be impossible to produce. Maybe you can estimate using national gas consumption?
edit: Found it, turns out: nope. https://www.fhwa.dot.gov/ohim/onh00/bar4.htm