The “unifying framework” that the US Bureau of Labor Statistics follows in designing the consumer price index, according to the agency’s handbook of methods, involves attempting to answer this question:
What is the cost, at this month’s market prices, of achieving the standard of living actually attained in the base period?
So, yes, it is a little weird that 24% of the latest CPI, the main US measure of inflation, consists not of market prices but of “the implicit rent that owner occupants would have to pay if they were renting their homes.”
This “owner’s equivalent rent” tends to make non-economists’ heads explode, generating frequent criticism from investors and others. But there’s no sign it’s going anywhere. From 1953 through 1982, the BLS used a different measure based mainly on the prices of new houses and monthly mortgage payments before abandoning it in the face of theory-based critiques from economists and practical concerns about how much volatility it added to the CPI. So, yes, owner’s equivalent rent is weird, but there doesn’t seem to be an obviously better alternative.
Some important questions are being raised at the moment, though, about how the BLS measures the rent from which owner’s equivalent rent is derived. Contrary to widely held belief, this is not done by asking homeowners how much they think their houses would rent for. That question is in fact asked but is used to determine the weighting that owner’s equivalent rent is given in CPI (the 24% mentioned above). The monthly changes that determine the inflation rate are estimated from changes in rents on similar dwellings, and those changes in rents are measured by asking thousands of American renters how much they’re paying. (The 7.4% of CPI that is actual rent is also measured by this survey.)
Does this truly represent market prices? That is, if you’re on a two-year lease, or you’re a long-term renter with a good relationship with your landlord, does the change (or lack of it) in your rent accurately reflect what’s going on with the cost of housing? Probably not, argued economists Brent W. Ambrose and Jiro Yoshida of Pennsylvania State University and N. Edward Coulson of the University of California at Irvine in a series of papers, the first of which appears to have begun circulating in 2012, and Adam Ozimek (now chief economist of the Economic Innovation Group, a Washington think tank) in a 2013 Temple University doctoral dissertation. Better to focus on new leases and thus measure what Ozimek called “marginal rents”:
Marginal rents reflect market turning points sooner, and show a larger post housing bubble decline in rents. In addition, marginal rents are shown to forecast overall inflation better than average rents.
The experience of the past couple of years has done a lot to reinforce this view. Zillow publishes a monthly rent index that measures changes in asking rents for apartments and houses (so do Apartment List and CoreLogic, but I use Zillow’s here because it’s available in seasonally adjusted form), and it shows rent inflation accelerating rapidly during the first eight months of last year and decelerating since, while CPI shelter inflation increased only slowly last year and has kept rising this year.
Last month, the BLS released a working paper by two of its economists and two at the Federal Reserve Bank of Cleveland that more or less endorsed this approach. The authors assembled their own rent indexes from BLS rent microdata and found that “rent inflation for new tenants leads the official BLS rent inflation by 4 quarters. As rent is the largest component of the consumer price index, this has implications for our understanding of aggregate inflation dynamics and guiding monetary policy.”
The most important of those implications would seem to be that the Federal Reserve’s policy-making committee was behind the curve when it started raising interest rates in March — a year after rents on new leases started exploding — and could end up late again in pivoting to easier monetary policy long after rents have started to fall.