A look at the downward trajectory of the U.S. hotel pipeline following the Great Recession offers some hints at how far the number of rooms in construction could fall as a result of COVID-19.
HENDERSONVILLE, Tennessee—In April 2020, the number of U.S. hotel rooms in construction stood at its highest level ever recorded, around 220,000 rooms. This compares to the prior peak in December 2007 when STR recorded the room count at 211,000 rooms.
(STR is the parent company of Hotel News Now.)
As the U.S. slid into a recession because of the Great Financial Crisis, the number of rooms in construction declined rapidly, by approximately 160,000 to just more than 50,000 rooms in May 2011.
With the pipeline at its peak again in 2020 and the U.S. economy in recession, it is not unreasonable to assume that the trajectory of hotel rooms in construction will mirror past behavior. Obviously, the decline in rooms in construction is a function of rooms exiting and entering the construction phase. When the number of rooms exiting is higher than the rooms entering, the total declines over time.
Projects that exit the in-construction phase normally open for business, but there is a very small number of projects that move back to the final planning stage. Projects that enter the in-construction phase normally move out of final planning. As was the case in 2008 and 2009, some developers—even though preparations had moved forward—likely decided that in the middle of a recession it was better to wait a few quarters until business conditions improved before starting vertical construction. It is also possible that some lenders—citing uncertainty in the overall economy—requested more equity from the owner which in turn may have delayed construction.
The trend of projects entering and leaving the in-construction phase can be studied when overlaying the respective totals on the number of rooms in construction.
The right-hand scale shows the number of total rooms as a line chart which decreases from 212,000 to 50,000. The bar charts show the rooms that entered and exited the in-construction phase, highlighting a clear difference between rooms entering construction through 2008 and rooms entering thereafter.
Not surprisingly, projects that had broken ground continued to open, basically at the same pace between 2008 and 2009. The count decreased only in 2010, which is a function of fewer projects starting construction.
Examining only the rooms added to in-construction, and summing the rooms by years, reveals that projects planned and financed during 2007, and basically “shovel-ready” in 2008, did break ground. The number of new projects in the in-construction phase then rapidly decreased in 2009 and 2010. It is worth noting that even in the long dark days of the Great Financial Crisis, projects were started every month.
Applying the same deceleration trajectory to the current pipeline of 220,000 rooms—a decrease of 160,000 rooms, a 76% decline—yields roughly 52,000 rooms in the summer of 2023. Let’s check back then and see if this comes to pass.
- Projects will continue to break ground, no matter what the macroeconomic situation is. All demand is local, as the hotel industry is a street-corner business, and there are always some street corners where developers feel supremely confident in their ability to succeed.
- In 2008, the number of rooms added to and exiting from the in-construction phase was basically the same, so the reaction to the negative macro numbers took a while.
- Because of the long planning and permitting time, decisions about postponing projects is impacting the pipeline for years to come.
- Lastly, this is a cyclical industry, and 13 years after the prior peak, the industry showed a new in-construction record. In other words, no matter the sentiment today, the pipeline will recover.
Kelsey Fenerty is a research analyst at STR. Jan Freitag is SVP of lodging insights at STR.
This article represents an interpretation of data collected by STR, parent company of HNN. Please feel free to comment or contact an editor with any questions or concerns.