It’s not often I hear a claim the BLS is overstating inflation, but it comes from two people I respect, so let’s investigate. 
I created the above chart from a Cleveland Fed data download of quarterly data and compared it to end of quarter data from the CPI (March, June, September, and December).
Rent is the CPI rent of primary residence and OER is Owners’ Equivalent Rent discussed below.
Cleveland Fed Synopsis
Rent measurement determines 32 percent of the CPI. Accurate rent measurement is therefore essential for accurate inflation measurement, but the CPI rent index often differs from alternative measures of rent inflation. Using repeat-rent inflation measures created from CPI microdata, we show that this discrepancy is largely explained by differences in rent growth for new tenants relative to all tenants. New-tenant rent inflation provides information about future all-tenant rent inflation, but the use of new-tenant rents is contraindicated in a cost-of-living index such as the CPI. Nevertheless, policymakers should integrate new-tenant inflation into inflation forecasts and monetary policy decisions.
Owners’ Equivalent Rent
OER stands for Owners’ Equivalent Rent. It is the price one would pay to rent one’s own house from oneself, unfurnished without utilities.
It is the single largest component of the CPI with a current weight 25.7 percent. Rent of primary residence has a weight of 7.6 percent.
Rent and OER track together in my lead chart as they should. They are both measures of rent.
Contrary to popular belief, OER is based on measured prices not by the question “If someone were to rent your home today, how much do you think it would rent for monthly, unfurnished and without utilities?”
Rather, that ridiculous question is used to help set weights, not set measured rent (see previous link).
Whether or not OER should be in the CPI is subject to debate.
I think actual home prices should be in the CPI. But home prices are not in the CPI because they are considered a capital expense.
My counter point is “So what? Inflation matters, not just consumer inflation.” The Fed has messed up multiple time with a myopic spotlight on “consumer inflation” instead of inflation and home prices are definitely part of the inflation picture.
With that background out of the way, let’s investigate the claim the Fed is overstating rent.
Professor Plum
Danielle DiMartino Booth
It is a bit premature to declare inflation is transitory, especially without a definition, but let’s postpone that idea for a bit and focus on rent.
The “New Tenant Repeat Index” name inaccurately mergers two ideas, a repeat index and a New Tenant Rent Index (NTR). The above chart shows the NTR, not repeats.
Let’s strip out OER to simply the discussion. OER closely tracks rent anyway, as it should, and as shown in the lead chart.
New Tenant Rent Index vs Rent of Primary Residence
Notice the enormous confidence interval for NTR, currently -0.73 to +6.20. The Cleveland Fed does not have much confidence in its own assessment this quarter and neither do I (reason explained below).
But the key point is that it is a huge mistake to think that new leases represent a better measure than rent of primary residence.
We need to compare CPI rent to existing leases because renewals make up the bulk of rentals.
Cleveland Fed All Tenant Repeat Rent Index
Unfortunately, ATRR data is only through the first quarter, that I can find.
Nonetheless, the chart pretty much smashes the idea that the BLS is wildly off and that 2.74 is the actual year-over-year market price.
Secondly, please note that CPI rent tracks above ATRR from 2017 through 2020. And CPI rent should track higher because CPI measure consists of new and existing leases.
If new leases are hotter than renewals, it would be logical for the CPI to be above ATRR.
I cannot explain why ATRR is ahead of Rent in 2021 and part of 2022, but one possibility is the BLS understated rent, not overstated it, for about a year.
What About Lags?
Most analysts tend to think there is a lag of about a year on rent prices. This makes a bit of sense since in month 1 of a new lease the CPI only picks up 1/12 of it as leases roll over.
However, leases are not evenly distributed. Most leases are in the months of May through September as High School graduates leave for jobs of school.
For multiple reasons, conventional wisdom on lag times is suspect. The above annotated chart shows huge differences in estimated lags. And to top it off, ATRR disagrees with all of them. It’s about 6 months as shown by my chart.
The median new lease is not a great measure because it does not even consider the number of rooms. They have alternate measures which are based on a Case-Shiller type methodology that will yield yet another measure of lags. And Apartment List does not seasonally-adjust its data making a mockery of the whole thing.
If you have a favorite lag period, you can find supporting evidence no matter what it is.
What About Methodology and “Actual” Prices?
Please consider this snip by Politano, the source of the above chart, from his report The Most Important Inflation Indicator Shows More Cooling Ahead
The Bureau of Labor Statistics’ (BLS) measures of shelter inflation are based on contract rents (that is, what existing renters are paying today) which gives them the most accurate possible picture of households’ real expenses.
Getting an accurate picture of the current drivers of rental inflation requires a different way of analyzing the data. That is where one of the most important new inflation indicators comes in—the New Tenant Rent Index (NTR, formerly the New Tenant Repeat Rent Index) created by BLS & Cleveland Fed.
Plus, it’s worth remembering that the most recent NTR data comes with significantly higher uncertainty bands due to the index’s methodological limitations. The two most recent quarters of data have much smaller sample sizes since the CPI’s housing survey measures the rental prices for units only once every six months, meaning a unit could get a new tenant in January but would not be surveyed and added to the sample until May. This makes readings for the most recent quarter subject to significant possible change—the initial NTR print for Q1 2023 suggested zero year-on-year price growth, but was revised upward as new data came in over the subsequent two quarters.
Inflation revised higher? Gee, who couldda thunk?
Actual Prices vs Actual Prices
I’ve written about the measure twice before, but to recap, they use the same underlying microdata as the official CPI shelter measures to look only at price developments among the subset of housing units where new tenants have signed leases, giving a much better read on where the housing market is now and where official CPI housing inflation is headed
Questions of the Day
Booth: “On market-based metrics (you know, ACTUAL PRICES) we’ve averaged 2% for all of 23. Note that even the Fed’s ‘market-based’ metrics are corrupted as they include Owner’s Equivalent Rent for shelter”.
Which one is actual, the BLS methodology or the Cleveland Fed methodology based on the same BLS methodology but more subject to wild sampling errors and revisions?
And which lag time is real?
Summation
New leases are a poor measure by which to judge overall trends in the price of rent.
The repeat lease index closely tracks BLS rent of primary residence and OER, which it should, because the methodologies are the same.
The difference between BLS rent and the Cleveland Fed repeat index likely reflects nothing more than new leases being in the BLS measure but not the repeat index. The implication is BLS imputations have at most a tiny impact on measures of rent. And even if they do, the Cleveland Fed and the BLS methodology is reportedly the same.
Assuming new leases have been running hot, a case can be made the BLS and/or Cleveland Fed understated rents for a period of time.
Regarding the sudden drop of the ATRR in the first quarter of 2023: The repeat rent data is six months stale and the sudden drop could be an outlier or sampling error.
As noted by Politano, “The two most recent quarters of data have much smaller sample sizes.”
CPI Rent
Rent of primary residence, the cost that best equates to the rent people pay, jumped 0.5 another percent in October.
People keep telling me rent is falling, I keep saying it isn’t (and the data proves it).
Rent of primary residence has gone up at least 0.4 percent for 27 consecutive months although the widely believed lag is 12 months!
For discussion, please see Falling Rent is Extremely Rare, Yet Economists Keep Expecting That
Booth and Professor Plum may be accurate about what’s ahead.
I am not knocking their viewpoints specifically because they could easily be right. Rather, I question the reasons they use to support their views.
They may be correct, but not because the BLS methodology is inaccurate vs the Cleveland Fed, not because of new tenant lease lags, and not because of theories about units under construction as widely believed and discussed below.
What About Housing Units Under Construction?
A widely-believed theory is that rents will come crashing down as soon as that massive block of housing units is completed.
I saw that theory so many times that I almost started believing it myself.
However, the data shows no discernable correlation no matter how you shift the lead or lag times.
The chart looks totally random. So, perhaps rents abate. Perhaps not. The data itself provides no reason to believe anything.
Meanwhile, please note year-over-year rent has a floor of about 2 percent except in the Great Recession housing crash. The average appears to be above 3 percent.
But if rent does drop, expect to hear “See, I told you so.”
Inflation Supporting Points
- The tail winds of global wage arbitrage and just-in-time manufacturing have reversed to head winds of onshoring and just-in-case manufacturing.
- Neither party will fix deficits and out of control spending.
- Trump’s tariffs and sanctions were hugely inflationary but Biden is much worse.
- Biden’s energy policy and regulatory madness is hugely inflationary.
- Retiring boomers need more medical care services. Their jobs are replaced by unskilled zoomers with a totally different work ethic.
- Massive wage increases in union contracts over a many year period and ongoing minimum wage hikes in many states.
Please bear in mind that for years I was one of the biggest deflationistas around. But many factors supporting lower interest rates and lower inflation have changed 180 degrees from tailwinds to headwinds.
And we had 50 years of disinflationary conditions culminating in an absurdly low 1.0 percent 30-year long bond yield.
Is it too much to think inflationary conditions and higher rates will last more than 18 months?
How Do Inflation Expectations Impact Wages?
As long as we are questioning conventional wisdom, now is a good time to review the idea that inflation expectations matter.
Please see my post How Do Inflation Expectations Impact Wages and Future Consumer Inflation?
Many will be shocked to learn that inflation expectations are irrelevant to future inflation. I can show this with data and explain why it is logically so.
What about the Inverted Double-Humped Yield Curve?
The yield curve went from steeply inverted to nearly flat and is now becoming more steeply inverted.
Double-Humped Inverted Yield Curve
I discuss the above chart in Huge Moves in the Yield Curve This Year, What’s Going On?
Regarding the huge inversion between 1 month and five years then strongly steepening: Could it be the bond market smells a short quick recession followed by a big inflation problem coming down the pike?
See above link for discussion.
Inflation may very well be “transitory” but for how long and to what?
I think we need a definition of “transitory”.
Footnote
On December 8th I received an email from the Cleveland Fed: “The paper was produced by researchers from both the Cleveland Fed and the U.S. Bureau of Labor Statistics (two of the four researchers were from the BLS). Can you update that in your article (and perhaps where it was syndicated)?“
Sure: Disentangling Rent Index Differences: Data, Methods, and Scope is by Brian Adams, Lara Loewenstein, Hugh Montag, and Randal J. Verbrugge. Adams and Montag are from the BLS. Loewenstein and Verbrugge are from the Cleveland Fed.
The Fed researchers could not answer any questions due to an FOMC blackout window. The Fed meets of December 13.
But they did say “Unfortunately, the ATRR index isn’t yet available beyond what you already have.”
Again, the Cleveland Fed methodology and BLS methodology is similar. Neither is more “actual”, and the ATRR data from the Cleveland Fed lags the BLS by nearly 6 months.
Mish I’m sorry about my recent outrageous off topic comments.
If BLS went back to the pre-1983 method of calculating housing inflation, then today’s housing inflation IS being underreporting in terms of its effect on headline CPI. Back then, housing inflation included previous sales price comparison like Case-Shiller uses
It’s that simple. The BLS made these changes in order to ensure housing didn’t over inflate CPI to the detriment of government spending on annual SS COLA increases.
So, yes, housing inflation by today’s sky-high prices is lowering CPI from where it really is. A first-time home buyer could tell you that.
How ’bout we just say the obscene Cost of Housing, intentional or otherwise, is likely here to stay.
At this point “fixing” it would require? result in? extreme upheaval of the greater so-called economy.
California has a statewide cap on rent increases. My rent will need to increase at 10%/yr for the next two years, just to catch up to *current* asking rents. Since California accounts for at least 13.6% of the naton’s total renter households, California alone will contrIbute 1%+ to the nationwide rent CPI for the next two years, assuming that all other rentals nationwide have no rent increases whatsoever for the next two years.
RE: Owner’s Equivalent Rent
Most homeowners are out of touch with the rental market so they will tend to underestimate how much their house would rent for.
Even if the BLS is not using the owner’s “guess” directly, the very fact that they factor that guess into their “weighting” helps to undercount inflation.
See also how they use “retained earnings” to estimate health insurance costs, which according to the official CPI are lower than in 2019.
Landlords : where will they go. If the answer is nowhere, bc the vacancy rate in the area is low and market prices are higher, greedy landlords can hit tenants with high increases.
Most renters don’t move. Rent increases in the flyover areas are higher than in the
major cities. The starting point was lower. Rent in the flyover is catching up, rising between 10%/20% a year. Most landlords expanded in the good years when rates were lower, seeking power and prestige.
$3T of business debt, mostly junk and 7T of gov debt will mature in the next year,
year and a half. Investors parked trillions in the gov roach motel to finance gov debt.
Next year, that wouldn’t be good enough. The Fed might suppress the long duration
and the middle, but junk debt rates will popup. Landlords who built a pyramid of multi
rental assets will have difficulties to rollover debt. They preempt by increasing rent,
hitting existing tenants…
I live in a small one stoplight town.
Talked to a local landlord who has a few houses and informed me rents have doubled in the last year here.
place has 75% owner occupied so those few few rents available get fought over. have a few places still for sale but their fixtures must be gold plated with what they are asking.
so even if you buy one and try to rent it the numbers don’t work, even at these crazy levels
Pshaw, that’s only a 2X increase.
That’s what the Fed intended all along.
Someone there mistakenly used a % symbol instead of an X.
I say that with high confidence that whoever came up with the idea of owner equivalent rent must have had a PhD in economics. Probably also jealous of peers getting Nobel prizes for similar quack theories.
Nothing at all changes hands, but lets measure it, when we can’t even measure house sales properly, that actually do take place.
Comment of the Day: “I say that with high confidence that whoever came up with the idea of owner equivalent rent must have had a PhD in economics. Probably also jealous of peers getting Nobel prizes for similar quack theories.”
It is stupid, but we have to deal with reality.
I am a strong proponent of abandoning the CPI and PCE and using a measure of inflation that uses actual home prices. No one ever asks me or even listens to me. Been writing about this since 2005!
“I am a strong proponent of abandoning the CPI and PCE and using a measure of inflation that uses actual home prices.”
Not only would you get a closer approximation of actual inflation by substituting actual home prices one-for-one for the current mumbo-jumbo. You’d get a better; much better even; approximation still, if you replaced the entire current so-called “inflation” calculation, by one consisting solely of actual home prices.
Inflation is personal. My basket may, or may not, be the basket the CPI uses. I can report that my basket says inflation is high enough to worry me for the next 20-25 of my life expectancy. I don’t see prices returning to 2020 prices and my retirement income is not keeping pace with prices. Good thing I have housing because the prices and interest rates make homes unaffordable.
Correct, inflation is personal especially to the 36% of renters and anyone who has to buy their own health insurance.
Calling Out Brian
Me: ““New leases are a poor measure by which to judge overall trends in the price of rent.”
Brian: “Dumbest thing I’ve heard today. New leases ARE the price of rent today. Over time, all lease rates converge on the new lease rate. Renewals tend to run at 50% or so of expiries, so at worst lease rents should be 50/50 new vs renewal rates. For inflation, it barely matters – renewal rates are calculated from new lease rates in most rent software. Professional MF real estate investor with $5B of real estate.”
Me: For starters only 16 percent of renters move each year. The figure is 5 percent for home owners.
Second, what is it about the data Brian fails to understand?
New leases means a brand new lease not a renewal. Based on the falling price of a new lease, for two years people have been telling me rent is declining or soon will. For 27 months they have been wrong.
was about to write “article packs a punch” and then saw Spencer’s take…so will just say “ditto what Spencer said”
Great stuff and lots of cross currents at the moment (in general, not just rents and real estate) with affordability issues becoming more of a focus.
Wow, that was a lot to digest. I think one should look at Canada. Canadian house prices went through the roof, and now rent is catching up.
There are strong headwinds against the idea that inflation is transitory. These trends are extremely strong:
These three macroeconomic factors virtually guarantee a higher inflationary environment in the future. Inflation may wane a bit, but the dice are loaded in favor of higher inflation.
Regarding inflation, the term ‘transitory’ may be correct, but to the average person it is baloney. Sure the inflation rate may have slowed or even stopped, but the pain incurred within the past 2 years remains until income can catch up to the inflation rate. Think of it this way….You are being tortured, with the pain level gradually increasing. The torturer no longer increases the pain level, but you are still in agony from the existing torture.
There was a story on local L.A. news several days ago, of a renters protest. Apparently the City is considering allowing rent increases of up to 6%.
Agree that calling inflation transitory is both simplistic and early. Clearly some was, some wasn’t.
“New leases are a poor measure by which to judge overall trends in the price of rent.”
Dumbest thing I’ve heard today. New leases ARE the price of rent today. Over time, all lease rates converge on the new lease rate. Renewals tend to run at 50% or so of expiries, so at worst lease rents should be 50/50 new vs renewal rates. For inflation, it barely matters – renewal rates are calculated from new lease rates in most rent software.
New leases means a brand new lease not a renewal. Based on the falling price of a new lease, for two years people have been telling me rent is declining or soon will. For 27 months they have been wrong.
I’ve read about this software. It basically advises landlords that it will be more beneficial to leave units empty rather than lowering rental rates to attract tenants.
The puzzling question is why does the Fed think that by raising rates to absurd levels will cure rent inflation. In fact, higher rates may stifle construction of new units and actually fuel inflation. Or is the idea that by making people lose their jobs they will eventually go live under a bridge and lower demand for rentals, when in fact demand for rentals may increase given the unaffordability of mortgages?
Is this incompetence, or is the Fed’s real goal the one to stifle wages increases for the benefit of corporations?
Good question and I have made that point about apartment construction before.
As for the Fed, it is best to think of them as highly trained idiots. They are trained to believe in economic models that do not work and they refuse to give up belief in those models.
For discussion, please see link to mishtalk.com
There are some very humorous comments on models in that post.
Finally, when stupidity is a possible answer, it is far more likely to be the correct reason than conspiracy ideas.
The FRBs goal is to sterilize home equity and increase duration for MBS. Higher rates achieves this, which is why I believe the bond market is wrong in its estimation for 100bps in cuts next year. If anything they will stop the Balance sheet runoff and keep rates where they are at, at least that is what they could do, I would rather see rates stay put and the FRB balance sheet below $6T
The only thing The Fed has ever done, is keep interest rates absurdly low.
That’s what printing money does: Increase supply of credit -> lower the price of credit. Economics doesn’t get simpler than that.
The correct, free market, efficient, economically enlightened hence non-absurrd interest rate; would be the one which resulted in fully convertible Gold being priced at $20/oz.
As to your concern: That sort of interest rate, would have within a few percent of appartments firesold by this afternoon. With every bank and lender out of business. Hence noone in a position to obtain a loan for any of the firesold appartments. Appartments which would hence sell for an absolute pittance. Plug “pittance”; for OER or whatever; into the “inflation” formula, and you’ll see how even just properly high rates would invariably lower “inflation.” Absurdly high ones, say ones resulting in $5/oz, would only strengthen that effect.
The Fed’s sole mandate is to transfer as much wealth as possible; from competent, produtive people; to incompetents close to The Fed. The Fed does that by way of debasing the currency, and then providing the above incompetents preferential access to the freshly printed money thus printed. So that those incompetents can then go out and outbid competents; who instead had to work for their purchasing power; for all that is valuable. Housing included. That is what The Fed does. And also ALL that The Fed does. Ditto every other, existing as well as possible, central bank does. There are no exception. None existing, none possible. Not here and now, nor at any possible time in any possible universe.
If anything, I believe inflation is vastly UNDER-STATED.
That is why I propose a new indicator: EEF.
Eaters Equivalent Food is the cost in today’s dollars of what you used to be able to afford to eat. This indicator is about to accelerate dramatically higher.