MSCI: CRE Shouldn’t Worry About AI Negative Impact

One of the natural worries in commercial real estate — especially office — with the onset of ever more powerful artificial intelligence software is that there would be fewer workers and, so, even less need for a lot of leased space.

Jim Costello, executive director of MSCI Research, has a different take:

“Hype over the negative impact of AI on commercial real estate is on the rise. To hear some commentary, we should be welcoming our new robot overlords in the hopes that they will deign to give us menial jobs since all knowledge-driven activity will be replaced by machines. Forget about debates over the return to office — such a worst-case scenario caused by AI could be far more devastating to office demand. There are other, more hopeful views on the future, however, which could spell out even more demand for commercial real estate as AI gains a foothold in the economy.”

Costello makes a reasonable argument that change often can’t be easily summarized as “X happens and Y as a result.” He’s essentially pointing to a system engineering view, where interactions are complex and that a significant enough change causes a realignment and new balance point.

“Data from the latest edition of the MSCI Real Estate Market Size report suggests that, with higher rates of economic productivity, even more commercial real estate is needed,” Costello wrote. A graph he includes “shows the relationship between output per worker and the value of commercial real estate in each country, analyzed for 2022.”

“The chart shows that as the output per worker increases, the value of commercial real estate in each country generally increases as well,” he adds. “With more output, income, along with the need for commercial real estate to service the flow of earnings into consumption and investment activity, grows.”

Costello does note that the type of commercial real estate might change. “Maybe there is a knock-on effect to the current turmoil in the office market that further constricts office demand as capital is brought in to replace labor. (And maybe there is more demand for data centers from such substitution.),” he said. “But the increase in economic activity will generate demand for other sorts of commercial real estate, with travel and consumer spending potentially generating more demand for hospitality and retail property from the wealth effect.”

However, while there is value in the view, it doesn’t mean that ultimately everything will be fine and dandy. The productivity in question is labor productivity and it means how many hours of labor it takes to product GDP. More productivity is more production per hour of labor. There is an inherent push to reduce the amount of work.

There has long been an argument, bordering on assumption, that new technology results in more jobs even as it might display certain types of workers. That depends on two assumptions: that the new number of jobs will be equal in quantity to the old number, and that people can easily move between different types of work.

Neither is necessarily the case. A company doesn’t automate work so it can keep the same number of employees, even if it talks about repositioning employees for more “meaningful” work. That may happen with some, but the entire strategy depends on having fewer workers.

From a CRE property owner’s view, more demand for one type of property instead of another might be fine. But CRE owners, operators, and investors are, in an odd way, similar to employees. They have experience and expertise in certain areas. Someone who had been trading in office buildings for decades might find the shift to working in data centers not necessarily so smooth. And growing power of computing could mean less total space, as the computers should be able to do far more than amount of space an equivalent number of employees would take up.

Costello might be right, but assuming everything will be fine is not necessarily wiser than assuming it will all crack up.