Earlier quoted context omitted.
> While building in SF proper would decrease rents there I'm not sure that's a given. Building lots of luxury housing in SF might actually decrease rent in Oakland and other parts of the Bay Area but not have an appreciable effect on rent in SF proper, at least in the short term. In fact, you could even argue that because of network effects and gentrification, building more luxury housing in SF would actually increas…
The point they make in the article is that the rich are going to move to SF anyway, but without available luxury housing they will have to buy up cheaper housing and have it remodeled. So you still get gentrification and network effects, but even higher prices for non-luxury housing than if you'd have just built more luxury housing in the first place.
That's what I'm actually unsure about. The article states it as a given without data or addressing regional effects. There are at least some well-to-do individuals who would purchase non-luxury housing in the North Bay, Berkeley, and the Peninsula rather than SF proper. You can obviously point to activity in places like the Mission and Bernal Heights, but at least some of the demand for unbuilt SF waterfront condos is going to go to the Oakland waterfront rather than something more inland. You can re-model your home, but you can't re-model your neighborhood.
Even if gentrification is a given though, the timing isn't. SF interest groups are pretty good at slowing things down (whether that's new development or a purchase and re-model). Presumably there's some difference in utility to the rich of buying luxury housing and buying non-luxury housing and remodeling it (otherwise there wouldn't be much demand for luxury housing). That's leverage the poor have in extracting some kind of negotiated benefit, like better public transit, education rights, re-distributive taxes, and so forth.