by Calculated Risk on 11/01/2021 10:21:00 AM
Monday, November 01, 2021
Construction Spending Decreased in September
From the Census Bureau reported that overall construction spending was "virtually unchanged":
Construction spending during September 2021 was estimated at a seasonally adjusted annual rate of $1,573.6 billion, 0.5 percent below the revised August estimate of $1,582.0 billion. The September figure is 7.8 percent above the September 2020 estimate of $1,459.3 billion.Both private and public spending decreased:
emphasis added
Spending on private construction was at a seasonally adjusted annual rate of $1,229.9 billion, 0.5 percent below the revised August estimate of $1,236.1 billion. ...Click on graph for larger image.
In September, the estimated seasonally adjusted annual rate of public construction spending was $343.7 billion, 0.7 percent below the revised August estimate of $345.9 billion.
This graph shows private residential and nonresidential construction spending, and public spending, since 1993. Note: nominal dollars, not inflation adjusted.
Residential spending is 14% above the bubble peak (in nominal terms - not adjusted for inflation).
Non-residential spending is 10% above the bubble era peak in January 2008 (nominal dollars), but has been weak recently.
Public construction spending is 6% above the peak in March 2009, but weak recently.
The second graph shows the year-over-year change in construction spending.
On a year-over-year basis, private residential construction spending is up 19.3%. Non-residential spending is down 0.5% year-over-year. Public spending is down 2.4% year-over-year.
Construction was considered an essential service during the early months of the pandemic in most areas, and did not decline sharply like many other sectors. However, some sectors of non-residential have been under pressure. For example, lodging is down 32.8% YoY, and office down 2.9% YoY.
This was below consensus expectations of a 0.4% increase in spending, and construction spending for the previous two months was revised down.