Please accept my apologies, I should clarify. All the data I've ever seen on induced demand in major US cities suggests very strongly that building units in maj

Please accept my apologies, I should clarify. All the data I've ever seen on induced demand in major US cities suggests very strongly that building units in major cities does not create demand for a greater number of units than are constructed. The last study I saw on the subject in San Francisco showed that for 100 units of newly constructed supply, about 85 would be consumed by extant demand, roughly 15 would be consumed by new migrants, and less than 10 new units of demand that would not have existed otherwise gets generated. This is admittedly not obvious, and generally looks like purely generated demand when viewed at a very coarse level. Incidentally, induced demand is an idea I often see advanced as an argument against constructing housing.