Crosswords Sudoku and Comics
Science

Four-Year-Old Research Paper on Remote Work Predicted Collapse of Office Real Estate

A new study published in the American Economic Review estimated New York City's office stock would lose 47 percent of its value, and foreclosures suggest that prediction is holding.

150 W 28th Street, New York, United States
150 W 28th Street, New York, United States      Empty Office Building Manhattan    Breather breather / Wikimedia Commons (CC0)
By Free News Press Editorial Team
Published August 31, 2026 at 1:34 PM PDT

When a draft paper called "Work from Home and the Office Real Estate Apocalypse" began circulating in 2022, it attracted attention across real estate circles and mainstream media. Four years later, according to Phys.org, its central predictions have largely played out.

The paper was co-authored by Stijn Van Nieuwerburgh, a professor of real estate at Columbia Business School, along with Arpit Gupta of NYU and Vrinda Mittal of the University of North Carolina. It introduced the concept of an "urban doom loop," a cycle in which falling office values reduce city tax revenues, forcing cuts to public services or tax increases, making downtowns less attractive, and putting further pressure on commercial real estate. The paper received coverage in The New York Times and on 60 Minutes.

The research, published in the journal American Economic Review, combined lease-level data from more than 100 U.S. office markets with a new asset-pricing model. The researchers found that between the end of 2019 and the end of 2023, annual lease revenue declined more than 15 percent nationwide. Companies leased less space and paid lower rents. Firms that adopted remote or hybrid work reduced their office footprints the most. Cities and industries with greater exposure to work-from-home policies saw steeper declines in both occupancy and rents.

Because office leases typically run for years, the researchers argued at the time that the full damage would appear gradually, only as leases expired and tenants downsized. Their model estimated that New York City's office stock would ultimately lose about 47 percent of its value. Nationwide losses, they projected, would approach $557 billion.

Van Nieuwerburgh says the core argument has held up. "Office buildings are defaulting on their mortgages, with staggering losses, literally every day," he said. "As I predicted four years ago, this is a slow-moving train wreck."

Long-term office mortgages maturing in a weaker leasing environment and a higher interest rate environment have left many distressed buildings unable to refinance. Foreclosures and discounted sales have followed. The hoped-for return to pre-pandemic office attendance never fully materialized.

The market has also split in ways the researchers anticipated. Trophy office buildings in prime locations continue to command high rents. Older, aging offices face rising vacancies and costly deferred maintenance with few tenants willing to fill them.

Van Nieuwerburgh does see some adaptation taking place. Some obsolete office buildings are being demolished. Others are being converted into housing. Those conversions have attracted policy attention in several major cities as a potential way to address both the commercial real estate problem and persistent housing shortages. Whether conversion activity can absorb enough distressed office supply to stabilize urban tax bases remains an open question, and one that researchers in the field continue to track as long-term leases on buildings signed before the pandemic continue to expire.

Notes on the occupants of the 45 "bouweries" or farms mentioned on Joan Vingboom's map of "Manatus", dated 1639
Appendices: A. Van Rensselaer manuscript, 1630 (photographic copy)--B. Vingboom map (photographic reproduction)--C. English translation of the Van Rensselaer manuscript.--D. Full-size copy
Notes on the occupants of the 45 "bouweries" or f…      Empty Office Building Manhattan    Van Winkle, Edward, 1878- Vinckeboons, Joan. cn Rensselaer, Kiliaen van, 1580?-1646 / Wikimedia Commons (Public domain)