Homebuilding surge no surprise

Homebuilding surge no surprise

Monday, February 10, 2020 from Floor Covering Weekly

Nationally, there were 1.44 million housing starts on a seasonally adjusted and annualized basis in the fourth quarter of last year. That was the best quarterly performance for homebuilding since the second quarter of 2007, when the market was heading into the housing crash. However, the end of the year can be a tricky time to decipher broader market trends. Unusual weather patterns can distort the season adjustment procedures to make production levels seem unusually high or low. Impending regulatory changes may encourage builders to start projects ahead of schedule to avoid these regulatory impacts.

None of these special conditions seem to be at play this time, as the upturn was very broad based. More volatile multifamily construction levels were indeed very strong, but not that far out of line with recent levels of activity. What really has been moving the needle recently is the strength in single-family construction. Production levels last quarter approached one million units when seasonally adjusted and annualized, the closest that we’ve been to that threshold in well over a decade.

Even though the strong fourth quarter caught most forecasters off guard, we shouldn’t have been surprised by the strength that we’re seeing. Given the levels of population growth, the industry has been undersupplying the market for many years now. In the 1980s, builders built almost 15 million new homes. In the 1990s, the level dropped a bit to under 14 million, but then popped up again to over 15 million in the 2000-2009 decade. This past decade, we couldn’t even get to 10 million homes started nationally, making it the worse decade for housing production in the post-war era.

Projections by the Joint Center for Housing Studies suggest the we should be building over 15 million new homes over the coming 10 years. While that would be a reach given what we’ve seen in recent years, it would only be a typical decade by historical levels. The basis of this estimate is much more than a “we’ve done this before, so we can do it again” approach. Their projection methodology consists of a detailed analysis of the three sources of demand for new homes: the formation of new households; demand for second homes, vacation homes, time-shares or other housing units that don’t have a permanent occupant; and, homes that are needed to replace existing units lost to the housing inventory by fires, floods, other natural disasters, or merely tear-downs to make way for a more expensive replacement.

Moving forward, household formation numbers are expected to begin to pick up speed. Current estimates are that we’ll see over 12 million net new households formed over this decade, and each of these net new households will by definition occupy a housing unit. Add to this figure homes that don’t currently have a permanent resident. The number of second homes and vacation homes has declined somewhat in recent years with the decline in household formations. Still, there is projected to be a 1.2 million unit increase in homes in this category. Finally, homes that need to be replaced due to losses to the housing inventory have been increasing in recent years due to more extreme weather conditions. In recent years, between 1.25 percent and 1.5 percent of all homes are lost to the stock each year, generating a need for about 1.7 million homes over the coming decade to replace these losses. Summing up these components, housing starts should total just over 15 million over the coming decade, or just over 1.5 million a year. That would put the 1.44 million that we saw in the fourth quarter just a bit below the typical expected levels.

Kermit Baker is the senior research fellow for the Joint Center of Housing Studies at Harvard University. He may be reached via e-mail at [email protected]

Leave a Reply

Your email address will not be published. Required fields are marked *

1