The American Enterprise Institute collects the timeliest estimates of the decline in average house prices since the Fed began another interest rates cycle in May. This is what they found:
"House prices peaked in June, ending a boom that began in 2012. Since then we have had [month-to-month] declines of 0.2%, 0.4%, & 0.5% in Jul., Aug., & Sep., respectively. Falling prices are not normally counted as rising inflation (except in the CPI).
This is critically important because everyone is citing the July, August and September Consumer Price Index (CPI) as the main reason why the Federal Reserve will continue rapidly raising interest rates until something breaks.
The CPI puts a huge 32.5% weight on Shelter (40% for Core CPI). When economists fret about rising inflation in "services" they effectively mean housing services. Medical services get only 6.9% weight in the CPI.
The graph shows that if we exclude Shelter from the CPI, the average increase of all other prices (including food and energy!) was BELOW ZERO in July and August. It was up just 0.2% in September because of a seasonal adjustment (the unadjusted change was negative).
CPI less shelter inflation was zero for the third quarter. Notoriously outdated estimates of rent are what pushed monthly CPI inflation up 0.16%, which is still only a 2% annual rate. Why wasn't 0-2% inflation for three months not headline news? Because the press keeps obsessing over the latest 12-month year to year change, which is 92% old news and extremely slow to change.
One month of inflation is too few, but 12 months (year-to-year) is too many - each new monthly average includes the previous 11 months. Three months near zero, on the other hand, is too many to just ignore.
Since house prices fell in the Third Quarter, according to the AEI, why did the cost of "Shelter" dominate all other prices in the otherwise zero-inflation Consumer Price Index in July, August and September?
Answer: Because the CPI uses rents to measure housing costs, and the estimated rental value of homes and condos relies on Bureau of Labor Statistics workers slowly collecting samples of urban rents (the CPI ignores non-urban areas).
As a recent Wall Street Journal article explained:
"The CPI’s rent component...is estimated based on ...rents raised months ago...Because rents rose strongly...last year, those increases are now feeding into the CPI...Dallas Fed [economists]...see a lag time of up to a year and a half [for CPI rent].
https://www.wsj.com/articles/climbing-housing-costs-could-prop-up-inflation-for-a-while-11663641828
The FOMC projects that if all goes well according to their technocratic central plan, PCE inflation would get down to 2-3% by the end of 2024 - not the end of 2022. They have no target for CPI inflation for many good reasons: It uses fixed weights as if people did not buy different things as prices change, it only covers urban metros (whose CPI's vary widely). and it assigns at least twice as high a weight to rents as the PCE inflation index. But shelter costs are extremely late to reflect falling house prices in both inflation measures.
By raising mortgage rates to 7%, the Fed has forced many prospective home buyers to rent, so "tightening" ironically raises rents in the CPI in the short run. And by crashing home construction, the Fed is greatly reducing the supply of house and apartments and thus raising home prices and rents in the future.
Meanwhile, the rent-dominated CPI gives the Fed an excuse to keep thinking it's astonishing success in raising mortgage rates and shrinking retirment savings is having little effect on housing or total inflation. Why? Because the CPI won't get around to recording the delayed effects of falling home prices on rents for at least a year.