Back at Modeled Behavior I've dribbled out over the past months my thesis that contrary to common perception, America is closer to a housing shortage than a housing surplus. The prices of houses rose to extremely high levels during the 2000s. However, total home building did not.
What was noticeable about that period was the fraction of homes that were site-built single family homes, rather than duplexes, apartment buildings or mobile homes. However, the total amount of homes built barely reached records. In absolute terms not many more homes were being built in 2005 than in the early 1980s when the population was smaller and immigration less of a force.
I still think the reasons for the switch to single family building are not fully appreciated. I've argued that whatever its flaws might have been, the subprime boom should be viewed as a technological innovation that allowed millions of households to switch out of the market for mutli-family homes and mobile homes and into the single family market. This drove both a switch in the type of construction and pushed up the price of existing single family homes.
Yet, even more important for understanding the current state of the economy is appreciating that while the increase in home building during the boom was not historic, the collapse in homebuilding has been. For several years now the United States has been building fewer homes than any single month in the 40 years proceeding.