[Chris2048]

> People of lower socioeconomic status drive less safe cars on less safe roads for longer commutes.

But can't you account for 'type of car', 'type of road', 'commute length' as direct variables pretty easily without dipping into social/economic backgrounds?

[kube-system]

The socioeconomics of the situation is why I'm questioning it, not what I think would be best measured.

Although it certainly isn't "easy" to measure all of this directly; there are thousands of that constitute the type of driving scenario that someone might engage in. Even just "type of road" isn't a single thing, it's hundreds of things.