[Panzer04]

Huh, interesting. I guess there's probably more accidents in urban areas, but thanks to lower speeds fewer fatalities.

I don't quite follow the second point - my presumption is that the chance of a bike hurting a pedestrian is lower than a car doing the same, and the chance of causing a fatality is, in general, reasonably low compared to getting hit by a car.

Stats would probably be hard to gather - there's probably quite a few bikes hitting pedestrians, but in all likelyhood many incidents go unreported if no one is injured.

[Retric]

A single bike accidents is less likely to cause serious injuries but statistically that’s offset by vastly more collisions.

IE the number of serious accidents depends on the number of accidents times the risk of each individual accident and bikes are far less segregated from pedestrians than cars.