Kinda, sometimes. But this field is also fraught with tons of "bad implementations of a good idea", making the problem worse, as well as intuitions that don't hold up to empirical studies.

Classic example is a neighborhood adds extra bright streetlamps and neighbors install extra-bright flood lights on their garages. This illuminates some areas, but the areas that are missed become even "darker" because the extra light in people's eyes ruins night vision. Hiding in the bushes with your black thieving skimask (or along the fence line, or just off the road) becomes even more effective, and everyone else just gets a bright streetlight shining into their bedroom window all night long.

There's a whole subculture of designers talking about what makes effective street lighting, but it basically comes down to less blue light aimed more downwards.

And the best crime deterrent is having a lively neighborhood with more eyes on the street in the Jacobs-ey fashion. I heard of one crime study that considered lighting and cameras and all those things but found the best predictor of low crime rates was how many dogs lived in the neighborhood (and therefore people outside walking their dogs).

On the good side of the spectrum, I've never experienced a city as dark as Tokyo at night, which is also one of the safest on both accounts...

I don't want to walk in dark even if there is literally no other person.

> Also simply driving safety. Driving with just headlights causes more accidents than on a well-lit street, since visibility is so much worse.

Accepting your points for convenience, wouldn't we get the same benefits with lights that only pointed downward? And if you further restrict to warmer colors (which interfere less with low-light sensitivity, also less diffraction reducing light pollution), with a sensible but low max intensity (again keeping the human eye more dark-sensitive, allowing better visibility into the non-illuminated spaces)?

Reducing nighttime lighting by 50% would make a huge difference in sky visibility.

Lighting affects property crime more than violent crime. It also depends on where the lights are, both in terms of neighborhood[1] and country[2].

On the other hand, lights dramatically affect how safe people feel[3] in an area, even if that feeling is not necessarily based on actual danger.

[1]: https://doi.org/10.3386/W25798

[2]: https://doi.org/10.1111/J.1745-9133.2004.TB00058.X

[3]: https://doi.org/10.1016/0169-2046(96)00311-8