There is no expectation that population growth would coincide with
decreasing sale prices. The premise is that new construction could allow population growth without a large long-term increase in price. And a still-growing population would expect to see at least a temporary price increase as new construction lags behind population growth.
It isn't even expected that the long-term prices would be exactly flat, because higher density construction costs somewhat more (e.g. taller buildings have to be made of steel rather than wood), but you want that rather than demand permanently outstripping constrained supply to set the ceiling on the long-term price increase.
> For example, rents in SF fell by 30% during the pandemic. And not as a result of new construction.
SF had negative population growth during the pandemic. Prices reflect supply and demand, so you can also get them to go down by reducing demand, but that isn't typically the one you want.
> And I'm sorry, I'm struggling to understand the logic in your post
Suppose you have a method of alchemy that can convert an ounce of silver into a pound of gold and an ounce of bronze into a pound of silver. If you do it, the price of each ounce of the metal you converted has gone up, because it's now the more valuable type. People then say "why are you making the metal cost more when we want it to cost less"?
But if you do it a lot, the overall market price of silver and gold come down, because every time you do it you're creating more and reducing their scarcity.
> In reality, rents are not a good indicator short-term.
In reality, everything always has a lot of confounders. For example, rents and new construction aren't independent variables -- when rents increase it will tend to cause a corresponding increase in new construction unless it's inhibited by something, and indeed high rent (i.e. demand in excess of supply) is the primary driver of new construction. So if you look for the places with new construction, of course they'll be the places with higher rent. But the conclusion shouldn't be that wet streets cause rain.