Measuring Rent Growth Is a Mess for Government Economists. A New Index Could Change That

Researchers at the US Bureau of Labor Statistics and the Federal Reserve Bank of Cleveland have been in search of a better understanding of inflation. Their inquiry into the rate of change of residential rents suggests that government calculations of inflation, in which shelter is 32% of CPI, may be significantly off. They suggest a new approach that might better balance costs of renewals with rents for new tenancies with the launch of a new index that captures only the leases of tenants that have recently moved.

According to researchers Brian Adams, Lara Loewenstein, Hugh Montag, and Randal Verbrugge, this has become a particularly thorny problem in measuring shelter costs. Available indexes vary wildly. As they point out in their paper, in the first quarter of 2022, “the Zillow Observed Rent Index (ZORI) and the marginal rent index (ACY MRI) reached an annualized 15 percent and 12 percent, respectively, while the official CPI for rent read 5.5 percent.”

In economics, as in comedy, timing is everything. The timing has been repeatedly raised as a question by critics of Federal Reserve interest rate policy, which is why the Federal Open Markets Committee slightly slowed the pace of increases, because it takes time for change to percolate up.

The difference is something the commercial real estate industry, and particularly the multifamily sector, has understood instinctively for decades. Asking rents tend to be higher than renewals, because a similar increase for an existing renter would be more likely experienced as sticker shock, potentially causing them to move. That triggers property owner expenses like repairs and rehab to a unit and then the time lag for getting a new tenant in. To significantly increase rents through an entire building take significant time. Rents for new tenants, while important in the industry, don’t give an accurate accounting of an entire rent roll.

But that’s not how the government has been looking at rents, which affect calculations of homeowner rent equivalents, and so overall a big impact on determining inflation. At a given time, according to this paper, the BLS housing survey is “fully representative of the rental housing stock in US cities” and is a random sampling that should show rent growth.

Using the microdata underlying all the measurements, they found that the differences between CPI measures of rent growth and other indices were due to measuring for all tenants versus only new.

Which to use depends on what one is trying to do, and for monetary policy, it’s not clear which is better. Use an approach with only new tenants and the results can overstate the immediate impact. A blended index better represents today, but one of the issues of measuring inflation is trying to understand where the economy is going, and as the Cleveland Fed notes, “Rent inflation for new tenants leads the official BLS rent inflation by four quarters.”

There’s no indication of what the BLS might do or how the FOMC would interpret this explanation in the context of its mandate to control prices in the long-term. Any change is likely to take some time, but one possible result could be a push to continue higher rates for longer, as rents more broadly would reflect the higher indexes within a year.