Typically, construction is a more volatile sector than the overall economy. It falls further in recessions, but bounces back stronger in expansion. New residential construction historically has been a great leading indicator with it’s declines starting ahead of full-blown recessions, and rebounding earlier as well. Commercial construction on the other hand tends to lag the cycle as the newly empty storefronts, warehouses, and office space from businesses closing in the recession need to fill back up in expansion before the cost of new construction pencils out and banks are willing to lend to these projects. Same goes for lodging which lags the overall cycle even more.
This cycle is different, however. Strong residential demand is keeping the overall construction industry afloat and in much better shape than in past cycles. This demand is due to higher-income households being less affected by the recession and trying to take advantage of record low interest rates. When combined with the strong, underlying demographics of Millennials aging into their 30s and 40s, the housing market is hot. This demand translates into increased economic activity and rising employment in construction, logging and wood products manufacturing, but also in related industries like finance, insurance, and real estate.
Statewide, all of the residential strength is in single family this year, which is largely where ownership demand shows up. New multifamily projects have cooled, although a pipeline of future projects remains.
Of course statewide trends mask some important patterns. The strength in single family activity is largely in the state’s secondary metro areas with Bend and Eugene seeing strong gains year-over-year, Medford and Salem being slightly positive year-over-year, and Albany with declines. Given we experienced a recession this year, this level of activity is very encouraging. The Portland suburbs are seeing somewhat less new single family activity, mostly driven by a drop in Washington County (I assume due to South Hillsboro getting built out but I am not entirely sure there). Even so, the drop off in single family in the Portland suburbs is not large, and still stronger than a couple years ago. Furthermore strength over in Clark County (not shown here) offsets about half of the suburban drop on this side of the river.
In terms of multifamily activity this year, the biggest change is a big drop in Multnomah which can account for nearly half of all statewide multifamily activity during a year. So a change there really impacts statewide figures. On the other hand, multifamily activity in the suburbs and across the secondary metros continues to pick up. Large increase in multifamily are seen in Clackamas, Deschutes, and Lane with smaller increases in Linn and Marion. There are noticeable drops in Jackson, Yamhill, and across the river in Clark.
Overall the construction outlook is solid, albeit mixed. Housing and related sectors are not expected to linger in recessionary limbo. The forecast calls for economic activity to remain at relatively high levels. Even so, expectations are for minimal growth over the next couple of years. This is a clear departure from past cycles where housing starts plunged and then rebounded. Homeownership demand should remain strong but the rest of the construction sector is likely to face weakening demand.
This is particularly true for commercial real estate, which remains a key risk to the medium-term outlook. The combination of business closures, less foot traffic, and demand for office space points toward less new construction activity in the years ahead. Leasing demand for retail, office, and leisure and hospitality will need to rebound considerably for vacancy rates to decline and rents increase enough to justify the cost of new construction.
The chart below looks at inflation-adjusted construction spending in Oregon, to try and get a better measure of the volume of activity and not just prices (revenue). Our office does not do an Oregon nonresidential forecast as we lack good, timely data on starts and spending. The available information we have from Census comes with a lag. However I am using national forecasts as a guide and how they may translate into local activity, hence the projection label.
In terms of the overall construction outlook, the strength in residential demand is expected to largely offset the weakness in nonresidential activity. In the chart above, it shows a 23% drop in nonresidential this year and next. This is large but not as big as the 45-50% drops in each of the past two cycles. Given so much of current nonresidential activity is based on backlog and not brand new projects, slowdowns in the planning of new projects today affect activity over the next few years. The key here remains how many vacancies arise and what the absorption rate looks like as the expansion gets underway.
In terms of public works, the outlook seems to have balanced risks. On the one hand public revenues are doing better than first feared, so money could be available for such projects. On the other hand, hard budget choices will need to be made and policymakers may choose to spending money in other areas of need more than on large construction projects. We shall see.
Lastly, note that we will come back to housing demand and population growth soon. Portland State is about to finalize their 2020 population estimates.