The people can't come from rural areas because they aren't there to begin with, or because they can't leave (e.g. because they're farmers and that's where the farms are, or they don't have the money). To get a large enough influx for the demand to overcome the increase in supply, it has to come from places that already have a lot of people, i.e. suburbs or other cities. Which is where housing costs are a problem.
> This seems like a variation of the "luxury apartments are actually good for affordable housing" argument. Which of course is popular on HN, but not particularly true.
[citation needed]
If you build more housing then people move into it. Those people used to live somewhere else. The place they used to live opens up for someone else.
If the new place is above-average that means everybody gets to move up -- new supply lowers cost of luxury apartments from $3000/month to $2500, which means the previous $2500 apartments fall to $2200, which means the previous $2200 apartments fall to $2000 etc.
On the one hand, this means the effect on $750 apartments is relatively small because some of the benefit of increased supply is also going to people with $1500 apartments and $2000 apartments. On the other hand, that's actually what we need. It's absurd to have a giant cliff between regulated "affordable housing" and market rate housing. People at the 50th percentile need to be able to afford housing too.
And affordable housing away from a high demand urban center is no use to anyone and not going to attract anyone if there aren’t good jobs there.
But where is the extra population coming from? If you're talking about any other place, it reduces housing costs in that place. If you're talking about a rise in birthrates, that's theoretically possible (though by no means guaranteed), but the timescale is so long that there is plenty of time to build even more housing by then.
> And affordable housing away from a high demand urban center is no use to anyone and not going to attract anyone if there aren’t good jobs there.
Where do you expect the people to be coming from? If it's another city, that's still an urban center with jobs. If it's the suburbs, people can still live there and work in the city as they do now.
And they can't be from the middle of farm country because there aren't enough people there to be a major source of demand, even before accounting for the fact that they would have to both want to move to the city and be able to afford to.
As for where people are moving from other than farm country, there are plenty of economically struggling cities, towns, and suburbs.
The relevant numbers are relative to the status quo, not relative to zero. If we are expecting rising housing costs due to rising population, a method of lowering hosting costs is productive even if it only reduces the rate of price growth, since the alternative would be for prices to be even higher.
The only problem is if the method can itself cause population growth, e.g. by short-term lower housing costs allowing more people to afford to start a family. But that's the thing we have plenty of time to stay ahead of by continuing to build more housing.
> As for where people are moving from other than farm country, there are plenty of economically struggling cities, towns, and suburbs.
Except that they're in the same position as farm country. There aren't that many people there anymore -- Detroit has already lost more than 60% of its population since 1950 -- and the people who are there can't afford the cost of living in an expensive urban center.
The people who move into the city center when new housing opens up are predominantly the people who used to live in that city's suburbs.
Your argument would also imply a strong case for "economically struggling cities, towns, and suburbs" to eliminate density restrictions and attract economy-boosting density-preferring people to locations where housing is already more affordable.
The problem with this theory is that it's zero sum. If density attracts people and prices there rise, those people will have come from some other place. Then the prices in that place will fall and we still serve the goal of giving people an affordable place to live.
You can't increase demand in every place at the same time. It has to come from somewhere. And if you increase density everywhere at once, it also can't move from everywhere to everywhere. If one city increases density it attracts new people from other cities. If every city does it there is no net migration because the benefits accrue everywhere, so the demand in each place stays the same and the greater supply lowers prices.