Air conditioning rental market won’t show up in CPI inflation data anytime soon

Private sector data sources show that monthly rent growth peaked in the summer of 2021, moderated through much of 2022 and turned negative in September. Meanwhile, the CPI shows rents accelerating month-on-month at the fastest pace in 30 years. Why are these measurements so different? The response reveals a baffling challenge to the US Federal Reserve’s efforts to control inflation.

Rent is key to the CPI, but the methodology involved means the Fed is perpetually looking back. The CPI housing category accounts for nearly one-third of the CPI’s weight. It has two major components, the rent and the “equivalent owner’s rent” or REL. A bit of history helps explain OER. When inflation last peaked in the 1970s and early 1980s, rapidly rising house prices played an important role. So the government made a big change in 1983: it removed house prices from the CPI. House prices are volatile, and volatility is expensive since many pension and benefit programs are indexed to the CPI. (Also, economists argued that houses were an investment, not a “consumer” good, and therefore did not belong in the CPI.)

To replace house prices, federal economists created the OER. (Many people now know OER for the infamous landlord survey question: “If someone were to rent your house today, how much do you think it would cost?” is used only for weighting purposes.) The main variable of the OER is in fact the rent, taken from a survey of tenants. The OER uses the same rent survey as used for the “main residence rent” measure. .

A key difference is that the CPI rent measure includes utilities, but utilities are removed for OER (and moved elsewhere in the CPI). This puzzling methodological difference seems to serve little other than cosmetic purposes to make the two measures of housing appear more different than they are. The idea of ​​including utilities with rent likely predates sub-metering, which in the modern era allows utility companies to track and bill individual tenants in a multi-family building.

Conclusion: A continuous survey of renters far exceeds its weight class, serving as the largest data set in the CPI’s largest category.

So why is this methodology only now becoming a big deal? IPC hosting methodologies have been in place for decades, but have received little attention for a few key reasons. First, real estate players and the media generally rely on private sector data providers rather than public rent data. Second, CPI rent trends were so regular that they were rarely reported on. From 1992 to 2021, the CPI’s “primary residence rent” measure averaged 2.8% year-over-year growth, never exceeded 4.3%, and never only briefly turned very slightly negative (bottom at -0.6%) in 2010.

Against this backdrop, it is shocking to see CPI housing costs soaring now. Housing rose 6.6% from September 2021 to September 2022, the largest increase since the early 1980s. Even more alarming is the recent acceleration in monthly trends, which peaked in 30 years of 0.75% in September. These strong monthly increases will remain in year-over-year CPI calculations until the same period next year.

This is where it gets even more interesting. While the CPI shows a surge in rents, many private sector measures show that levels of rent growth have declined significantly in recent months. Data from RealPage shows that effective asking rents fell 0.2% in September, the first drop since December 2020. This dataset also shows that the largest monthly increases occurred in the summer of 2021, an increase which is just beginning to show up in the CPI.

So what’s the problem ? Why is the CPI rent survey so out of step with private sector data providers?

The two sources are measuring different things, a nuance that few outside of economists and Fed watchers appreciate. The CPI attempts to measure contract rents, also known as in-place rents. These are the rents a household pays if they stay in the same dwelling. Most private sector sources only measure asking rent for a new lease.

These methodological differences mean that the CPI is still significantly lagged. Some economists estimate this lag at around 12 months. This is also why CPI housing costs have remained more stable in 2021, even as new rents have increased. And this is also why the CPI did not take into account the larger rent reductions that occurred in 2008-2009 and, to a lesser extent, in 2020.

It makes sense that the CPI measures contract rents instead of asking rents, since contract rents capture what many households actually pay in rent. But it also creates problems. Asking rents are the leading indicator. Renewals are re-priced as leases expire, but are generally priced lower than new leases. A “reverse rent” scenario encourages tenants to move out and move back to another unit in the same community, which costs the landlord more than keeping that tenant in place. Once a lease is signed, this price is locked in for the duration of the lease – usually around a year – and will remain within the CPI metric.

This makes rent a particularly sticky component of the CPI, distinctly different from other goods like food or gasoline, which are both re-consumed (and therefore revisited) far more often.

So if you want to see where the rents are going, you start by asking for the rents.

The Fed is of course aware of this. But that still leaves them with a confusing problem. It looks like this.

1. For the CPI to cool significantly, the shelter component must cool.

2. Housing is most dependent on a rent survey, so rents need to cool down.

3. Any attempt to slow down rents will first appear asking for rents as a leading indicator.

4. Demand for rental growth is indeed slowing, but this will only show up in the CPI in 2023.

In other words: don’t expect the current slowdown in housing demand and prices (rentals and homes for sale) to impact the official inflation gauge anytime soon.

Guest comments like this are written by writers outside of Barron’s and MarketWatch newsroom. They reflect the views and opinions of the authors. Submit comment proposals and other comments to [email protected].