How big will the building boom be?

Here is a very simple model for thinking about housing production. It can be very simple because the dislocations are so large that very broad estimates get us into the ballpark, which is far away from where we are.

Figure 1 is an estimate (again, very simple and broad) of the number of new home completions that would have been required to keep housing at 20th affordability norms.

Figure 2 is the same chart, but in number of units instead of as a percentage growth rate.

This adds up to a shortage of about 15 million units, which is on the low side of what I think we need.

Figure 3 assumes that all the cyclical variations we have had remained the same, but it adds a valuation effect to reflect the added supply. I have used an elasticity of 4 (1% more homes = 4% lower prices) which is high. This is conservative in terms of estimating how many homes are needed to reattain affordability.

Figure 4 extends this into the future. I have charted a course here where in about 20 years the shortfall is made up and affordability is back to 20th century norms. I have lowered the neutral completions rate to reflect lower population growth.

One could debate my neutral growth estimate. I will lower it when I see firm confirmation that completions are high enough to stop the excess rent inflation. I don’t think that’s happened yet, so I think current production is below neutral. You can lower the rate a couple tenths of a point to suit your expectations.

Wherever neutral is, there will need to be a make-up period where we build the houses we should have built over the last 15 years, then construction will settle back down near the current rate of production or slightly lower, depending on future population trends.

The rate of construction during the make-up period will be pretty moderate in terms of the growth rate, compared to the pre-1990s period.

Most things grow exponentially. Housing production has not because it has been obstructed in one way or another for 40 years. In fact, it hasn’t grown at all.

This leads to a bias in our mental benchmarking. We have this nagging idea that there is a natural 2 million unit cap on the housing we need and the housing we are capable of building. There is no cap.

Figure 5, below, shows the number of completions in thousands of units instead of as a percentage of the stock.