Oregon Office of Economic Analysis

Housing currently faces somewhat contradictory data points. On one hand, housing demand is strong as evidenced by the home sales data and a declining rental vacancy rate. On the other hand, estimates of population growth range from modest at best, slowing more generally, to even small losses, depending upon the region or the data source. Regardless, it is clear there has been no pandemic migration boom in Oregon.

I see three main ways to square these seemingly contradictory data points. First would be the population estimates are too low, or the lagged data has yet to catch up to reality. Second would be the flow of new construction coming on to the market has slowed more than population growth has. Third is an increase in household formation rates among existing residents, which is the most interesting possibility.

We know that household formation rates have declined in recent decades. A small reversal of these trends would boost housing demand more than enough to offset weaker population gains. Such an outcome is possible given the large Millennial cohort is fully aging into their 30s this decade when living with roommates is less common. This can be seen in the chart below of headship rates by age in the Portland region. Headship rates are the share of the population that is a householder (formerly head of household).

There are two important items of note here.

First, headship rates increase significantly in ones 20s and 30s. This is the typical life cycle effect. The increasingly-middle-aged Millennials are right in the heart of these age cohorts. As such, Millennials are now the key demographic and overall driver of the economy. This is the structural, demographic tailwind for housing we have discussed before.

Second, headship rates among 20- and 30-somethings ticked up in recent years. The difference between the light blue and dark blue lines in the chart above may not seem like much at first glance. However it is actually a very substantial change. Among 20-39 years, this increase in just the headship rates is the equivalent to adding about 15,000 to 20,000 more households in the Portland region. That is a huge increase equal to more than a full year’s worth of new housing construction in the metro area. Now, some of this could be in part due to noisy data as the estimates do vary depending upon the exact years of comparison. In particular the 2020 ACS had a low response rate and is based on “experimental estimates” as opposed to official statistics. The good news is we are a handful of months away from the 2021 ACS data being released which will provide another snapshot of headship rates, and one which will be mid-pandemic. The main point here is not to get caught up in the specific calculation but to see how an increase in household formation rates is more than enough to offset slower population gains from a housing market perspective.

A key question is why would household formation rates increase? In the big picture we are probably talking about the same issues researchers were studying pre-pandemic on why household formation was lower. This included worse housing affordability forcing people to live at home longer or with roommates, in addition to things like delayed marriage or having kids, and so on. Changes to these big picture trends tend to be clear in hindsight, after a few years of data confirm any shifts.

So for now we are probably talking more about pandemic-related impacts to household formation. With a deadly, contagious virus going around, people likely wanted more space and to be around fewer people. Recovery rebates, record low mortgage rates, and working from home increased the demand for homeownership. At the same time new rental properties kept coming on the market initially as they were under construction in the years leading up to the pandemic. That meant there were more vacant units to move into. Combining all the above with the underlying demographics and it is not hard to see how household formation could increase in a meaningful way.

Over the longer-term, housing demand will come more from the rebound in migration patterns to Portland and Oregon. Migration is pro-cyclical. People move in search of better opportunities during good economic times. More tangible is the strong, recent rebound in the number of surrendered driver licenses at Oregon DMVs. Housing demand today is strong for structural, demographic reasons and cyclical, migration and economic reasons. That said, with affordability worsening again, will we see household formation slow as well, with fewer single-person households and more roommates out of financial necessity?