You can't just wave away economics like that. The problem with restaurants is that they are in a highly crowded and competitive business. If you give them a tax break, they would simply pass the savings onto the consumer.

Yeah. I think the issue is that the landlord is incentivized to extract maximum cash for the spot, which may be better for some other kind of venture (eg bank branch or whatever) which makes the restaurant's existence close to untenable in some locations. Public good or local quality or whatever does not factor in in any way.

There are probably better ways to do this than what I suggested.

Even if the landlords charged no rent, I would not expect to see profits increase significantly, rather I would expect prices to lower (and additional restaurants to open). Of course, if only a small(ish) number of restaurants get free rent, their profits would go up.

Fair enough. Consider my problem statement (rent extraction) rather than my poor solutions.

Or the restaurant owns the building.

Yeah, maybe not practical in SF, but elsewhere it works well.

Or long-term lease.

Typically NY situation is, up-and-coming restaurateur opens in a crappy underutilized space, takes a long-term e.g. 7-year lease. Lease is up, every established restaurateur and dilettante thinks they can make a fortune there, rent rises to where nobody can make much of a buck, often new restaurant closes shortly after.

I don't know why there are so many freaking banks, New Yorkers refuse to walk and are worth money to banks. But a restaurant is a huge pain in the ass to a landlord compared to a bank.

I wonder what the effects would be of cities requiring a long term lease to get a restaurant permit? Would landlords just not go along and we'd see fewer restaurants? Maybe, but it could be an interesting experiment to find out.

Zone banks to second floors (external stairway?)

Yes, that works too.

I remember when I was living in NYC when all four corners at an intersection went from restaurants to bank branches. Just awful.