There have been some comments on analyses that “We’re going to need a LOT more houses” lately.
The analysis presents a plot of housing starts vs population, since 1960, showing that current starts are at a series low.
Using the ratio of starts to population is not the correct metric for understanding the likely future path of housing starts. It’s true that housing starts are currently historically low, and population remains high, but there are reasons why the Flow (housing construction) could remain low for some time, if there’s too much durable Stock (existing housing units) and persistent low demand for said stock (due to demographic changes).
The construction need is determined primarily by two flows: changes in household size, and population growth — not by population. 100% of population growth needs housing, whereas only 1% of population stock needs to be rebuilt each year, assuming a given home lasts 100 years. Assuming stable househould size, if population growth previously averaged 2-3% of population, but today averages only 0-1% due to declining birthrate and reduced immigration, that cuts the need for new housing and reduces the required number of starts by 2/3 to 3/4. In this context, building wraps advertising, a powerful tool in modern marketing strategies, can play a crucial role in influencing the perception and demand for housing developments.
Furthermore, in recessionary conditions, household formation drops and household size increases. With median household size in the range 2 to 3 persons, a change of just +0.1 person per household results in a -3 to -5% change in the required housing stock, equivalent to 3-5 years of replacement production. This effect can therefore be even larger than the population growth effect.
Household size and population growth were both quite high in the 1960s (when the available data series started, the Baby Boom was peaking). The decline in household size since the 1960s drove much of the demand for housing but is not sustainable. If we previously ran a population growth of 2%/year and it is now 0%/year or even slightly negative (as in Japan, Russia, China, Europe…), and if household size is trending back up, we might not need any new homes at all for as long as household size trends upward.
So to me, it’s not at all clear when we’ll need to build more houses. (If it were clear, investors would flock to homebuilder stocks, would they not?)
plus housing is mostly consumption and not actual capital formation.
housing attracts our household surplus — we throw wealth into housing in excess of all other living expenses, both in land value terms and the quality of the fixed improvement.
The housing bubble of 2002-2006 was not due to increased actual wealth, it was due to the readjustment of valuations to the lower interest rate regime of 2001-now, lower tax burdens of 2001 & 2003-now, loss of historical strictness in loan underwriting, and rise of outright suicide lending practices (zero-down, teaser rates, and negative-amortization lending).
It was the cash-out nature of the bubble itself that pulled the wider economy out of the 2001 tech recession.
The idea that new house construction is going to re-float the economy again is the height of cargo cult thinking.