NEW YORK – How come the shelter component of the US Consumer Price Index is rising at just 2% a year while all the private-sector gauges of rent inflation show rent hikes of close to 10%?

Shown in the Chart of the Day is a comparison of the Zillow Rent Index with the Consumer Price Index rent component. Shelter is two-fifths of the total index, so the question of whether inflation is “transitory” or not depends on rents.

This question has a simple answer: It takes the Bureau of Labor Statistics about a year to catch up with rent inflation. What the government reported as rent inflation in July reflects the economy of a year ago and more.

That’s because renters’ leases take a year or two to expire. The Zillow, and CoreLogic reports reflect the average rent on a new lease, not the rent on leases signed in the past. As old leases expire and renters have to pay the higher market rate, rent inflation will surge.

The Fed can continue its Groucho Marx impression (“Who are you going to believe –me, or your own eyes?”) only until the inflation now in the pipeline shows up in the official numbers. Perhaps America’s monetary officials will find themselves at Dulles Airport this time next year surrounded by murderous mobs of renters waiting for rescue flights to Davos.

But the blame for all this will land on the desk of beleaguered President Biden, who has to answer to Americans who have seen their real incomes eroded by inflation during every one of the past twelve months.

The chart below shows that today’s reading of the Consumer Price Index for shelter lines up with past changes in the Zillow Rent Index. That is, past values of the Zillow Index correlate closely with today’s value of the Consumer Price Index, for the reasons I cited above.

The table shows the correlation between, respectively, the rent component of CPI and the Zillow Rent Index (in terms of year-on-year change), with lags. So the first bar (starting from the bottom and going up) is the correlation of this month’s value of each variable with the previous month’s; the next bar is the correlation of this month’s value with the value two months ago, etc.

What this cross-correlogram shows is that there is a very high correlation between this month’s CPI shelter number, and lagged values of the Zillow index going back 12 months.

Using the lagged relationship between the shelter component of CPI and the Zillow Index, I forecast next year’s change in the CPI shelter index from this year’s change in the Zillow Index. The result is shown in the chart below:

The model forecast lines up quite well with ten years of history of the CPI shelter component. Projected into the middle of 2022, the model forecasts a 5% annual rate of increase in CPI shelter. That’s far in excess of the Federal Reserve’s threshold of pain.