Housing: Part 183 - A measure of the shortage

Here is a graph of housing units per adult in the US, quarterly.  Unfortunately, there is a sizeable discontinuity in 2002 where the count of housing units was adjusted as a result of the decennial census.  But, even with that discontinuity, the trends are sharp enough that we can see pretty clearly that housing units peaked in the late 1980s and have been generally declining since then.

During the big, bad 2000s, we just barely started to reverse the trend before we busted the housing market in 2007.  (That little bump up in units per adult happened in 2008 after housing starts had collapsed, and is partly related to a slowdown in population growth.)  So, when we had a moral panic about housing, when we were supposedly overbuilding, and when capital was supposedly being unsustainably misallocated into housing, we were near secular low points in units/adult, hovering, as far as we can tell, around the level last seen in 1986.

Of course, the misallocation didn't come from where capital was allocated.  It came from where it couldn't be allocated - the Closed Access cities.

In any case, we can use this ratio as a first estimate of how many housing units we are short of some stable level that might have been built if we hadn't had the moral panic.

Oops.  Obviously, that y-axis is in thousands, not millions.

Next, is a graph of the number of housing units that would be required to just get us back to the same ratio of units/adult at the end of 2006, or alternatively, the number of units required to get us back to the nadir of the housing stock in 1991.  We need somewhere around 5 million housing units.

Eventually, as the baby boomers age into very old age groups, there might be some decline in housing units/adult, but we aren't anywhere near that now.  I don't think there is any current demographic trend that would lead us to expect the demand for housing units/adult to have shifted materially, as it did in the 1970s and 1980s when household size declined for a number of reasons.

This basically matches the gap suggested by housing starts.  Adding about 1/2 million units per year since the bust would put housing starts at near historical norms, and would roughly have given us those 5 million units.

How many hundreds of thousands of men without college degrees have been unemployed for the past decade while we patted ourselves on the backs for making this happen?  But we blame trade with China, because construction employment was unsustainable.  Right?  0.55 housing units per adult was just masking the loss of manufacturing jobs.  Right?  We had to do this.  It was inevitable.  Necessary, really.  Right?