So we end up with a pretty ugly societal tradeoff if we only look at this as some sort of voluntary choice, when it is not really so.
The interesting question (to me at least) is how, and how much, do we optimize our redistribution to help less well-off folks?
I think you need to take a whole-system approach here. A dollar spent improving air quality next to this power plant might be a great investment, or it might be way down the list vs. say funding universal healthcare, which would help with other (potentially more serious) health issues too.
QUALYs are a good tool for assessing the impact of a dollar of spend as it pertains to heath outcomes, but they are not the only lens. But until you have unlimited dollars, you need some framework for allocating them (after first deciding how much redistribution should be done in the first place).
In general, I think more choices are better than fewer, so having the choice to in a cheaper but more noisy neighborhood is probably better than being priced out entirely and having to move to rural CA where you can afford rent. (At least, revealed preferences show that most residents think that it’s the preferable option).
Problem: the only food poor people can afford is wonderbread. Solution: only accredited investors may now buy wonderbread.
If you only make it illegal to build housing near pre-existing industry, then what? That's basically a let them eat cake argument (not saying that that's what you're suggesting), because now there aren't cheap options near industry, and unless zoning changes allow for much more housing in the non-industrial areas, then the supply of housing goes down, the price goes up, and poor people just have to deal (all in the name of being saved from industry).
https://www.huduser.gov/portal/pdredge/pdr_edge_featd_articl...
Americans love their class symbols, and chief among them neighborhood. It's therefore absolutely necessarily that public housing be sprinkled around evenly.
Meanwhile, in the Austrian example given in the link above, Singaporean home subsidization, and other successful cases what you find are schemes that manage to create and maintain an admixture of incomes. There are different ways[1] to accomplish that and they're not all easy, but it does seem like a necessary requirement for success.
[1] As in the Austrian example, you can remove income requirements after move-in, so that people aren't forced out as they move up the income ladder. Or rather than subsidize rent you can subsidize ownership. By having an ownership stake people are incentivized to maintain the property, and even if they move they'll be replaced by someone with a similar income (or at least an income above the initial ceiling). Unfortunately, many liberals who promote subsidized housing scoff at permitting people to continue to benefit from that initial subsidization after their income has increased. But that resistance is precisely the kind of policy dynamic that concentrates poverty!