🚨🚨🚨 @uscensusbureau is making a mistaken decision to degrade the extraordinary high frequency labor market data they provide in a way that will both make it harder to track the economy in real time and to do economic research. census.gov/content/dam/Ce…
Not to be overly nationalistic, but thanks to @uscensusbureau US produces monthly micro data on the labor market that is vastly superior to what most or perhaps any other countries produce. Countless analysts rely on it to understand what is going on. This would make that harder.
This issue was flagged by @nick_bunker who pointed to, among other losses, the Atlanta Fed wage tracker & other analysis that adjusts for the changing composition of workers (we used to do this @WhiteHouseCEA & @arindube has been doing a lot of it lately).
The decline in retail sales might actually be a positive sign of a healthy normalization.
The economy has struggled to produce enough goods to match the voracious consumer appetite for them. If goods spending continues to normalize that would be, well, good.
To be clear, the chart in the previous tweet was real retail sales. Nominal retail spending remains very high but in it is no longer buying massive amounts more stuff than it used to be.
(Retail sales is mostly goods but does include services like restaurants and bars.)
Got new data directly from the BLS! They put together an estimate of U.S. inflation using the European concept for core HICP that is (hopefully) reasonably comparable to what Europe publishes. Here is the picture of the annualized 24-month changes compared.
The only published data is for the U.S. overall HICP (you can find it on FRED) but BLS does not publish a core series. There are still various issues with comparability, both in the definition of core and how they're constructed, but they're close. fred.stlouisfed.org/series/CP0000U…
Previously I had been using U.S. core inflation excluding shelter as my rough and ready guess of the most comparable to the Euro area core HICP concept. That guess looks pretty reasonable when compared to the new series BLS produced, at least recently.
All of these four factors are reasons that inflation in 2022 could be even higher than it was in 2021, something I would put about 20% probability on (speaking about core inflation, excluding volatile food and energy).
In the base case they would still be outweighed by transitory supply chain issues in used cars and demand shifted to goods.
Bottom line: guess of 3-4% inflation in 2022 (towards bottom of range for PCE and top of range for CPI). Wouldn't be surprised by 1% or 6%.
A closer look at inflation in services excluding energy, this accounts for about 60% of the CPI and 75% of the core CPI.
Over the last three months up at a 4.6% annual rate compared to 2.9% annual rate pre-pandemic.
That excess is adding ~1.2pp to annualized core CPI.
Overall core services are still below trend--but is moving up more quickly so converging towards/above trend. You could look at that and say that our only inflation problem is goods. Or you could look at that and say we have another shoe yet to drop.
Leading candidate for yet-to-drop shoe #1 is housing prices. These are still up relatively slowly for *all* housing as compared to the large increases we're seeing for new leases. All should catch up to new.
Headline CPI was 0.5% for December, core was 0.6%. Cars were a big part of the number (again) but inflation continues to broaden--the "other" excluding cars and pandemic services is high for the third month in a row.
Inflation is still almost entirely driven by durable goods not services. Durable goods inflation should come down as supply chains unsnarl but what will happen to services is the big question--is drifting up a little bit lately.
One reason to expect services to rise more is that they include shelter--which includes rent and owner's equivalent rent. The CPI is showing a much smaller increase than other measures. They're not comparable but measures of new leases show the future for all leases.
My talk: "Why Did (Almost) Nobody See It Coming and What Does That Mean for What’s Next?" from an ASSA inflation session w/ @JonSteinsson@GagnonMacro@jhausma1.
"Why did nobody see it coming" was stolen from the Queen's question after the financial crisis.
In some ways it is less important since inflation is much less bad than the global financial crisis.
In other ways it is more important: inflation *should* be more predictable.
I think we know (almost) everyone missed it. More than a third of the way through the yr private-sector forecasters thought 2.3% and put the odds of 4%+ at 0.5% (it ended up 4.5%). FOMC just as bad. And TIPS breakevens missed the CPI by even more than either of them.