Housing starts rose last year for the sixth consecutive year, bringing new construction up to 1.1 million units. Will construction get back to historical good times, which many think of as 1.6 million starts? That's very unlikely, because our population growth simply does not justify much more construction than we currently have.
Expectations of normal were formed prior to the last recession, when starts averaged over 1.5 million units. If one dropped recessions, the average was 1.6 million. Of course, if you take any series of numbers and remove the low ones, the average is higher.
For a reasonable expectation of construction in the coming years, start with population growth. The United States population increased by 2.5 million people last year, according to the Census Bureau. We live about 2.54 people per household, so population growth justifies about one million new housing units.
Reality is more flexible than this calculation. Young adults moved back in with their parents during the recession, reducing the need for housing units. Then last year many of them moved out, increasing housing demand beyond that indicated by population numbers. In addition to young people living with parents or not, singles may have roommates or not, and some married couples get divorced when they can afford it. So in any particular year, additional housing demand can be higher or lower than dictated by population trends. Over time, however, population drives housing demand.
(The data on households and occupied housing is soft; for more details, see Bill McBride's comments on the subject.)
Population growth has slowed remarkably in the past decade. The United States added 3.5 million people in 1992, a far cry from our recent 2.5 million growth. To generate a more optimistic forecast of housing, try boosting population growth projections.
There will be a population increase as the Millennials enter their prime child-bearing years, but the big swing factor is immigration. In the recession, we actually had more Mexicans leaving the United States than entering. When they lost their jobs in factories and construction projects, they returned home. I find it unlikely that immigration from Mexico and Central America will return to past levels, as their economies have improved and the U.S. has bolstered border patrols. And if Donald Trump is elected, look for that giant fence.
The near-term outlook is favorable, thanks to job growth that enables young people to live away from their parents and without roommates, as well as enabling divorce. Low mortgage rates help, though tight rental markets limit housing demand. The current economic expansion can drive housing construction above a million units for a year or two, but at some point slack population growth will prevent further increases in residential construction. That's likely to coincide with rising interest rates, double-whammying the housing market along about 2018, give or take a year.
The key takeaway for contractors and companies in the building products industry is that historical norms don't apply in a world with light population growth.