[Fradow]

My understanding is that the author is arguing that:

- guardrails should always be between the sidewalk and the road. Not after the sidewalk

- in places where statistical data shows collision or where there's a high risk of cars going on the sidewalk, bollards should be installed. A prime example is in parking lots where cars park facing the sidewalk.

[aidenn0]

Anyone know if we can reasonably estimate X for "A bollard installed here will save X/100000 lives per year on average" for various spots one might want to install bollards, and what the CI would be of such an estimate?

Presumably in cities with enough traffic it's possible to empirically measure number of times a car jumps the curb per year, but in other areas maybe not?