gentlemen. steven corbett Assert that the Philadelphia Fed’s early preliminary benchmarks support a recession in the first half of 2022, namely:
You, Menzie, are more likely to be correct with the Est Survey. You wrote: So: (1) I give more weight to the institutional series, (2) the gap between the two series is more likely to be due to increasing and biased measurement error in the family series, rather than, for example, being primarily There has been an increase in the number of people wearing multiple hats. https://econbrowser.com/archives/2022/12/the-household-establishment-job-creation-conundrum
Big mistake, because it turns around. As expected.
You’re wrong because you haven’t considered statistics more holistically. This is where your students learn. If your dial is telling you different things, cross check your indicators. If employment is growing faster and faster, then GDP should also be rising. If jobs are growing rapidly, so should mobility and gasoline consumption, because so many people in this country drive to work. Finally, if productivity is collapsing while jobs are being added, you really need to pause and put together some narrative about why this is happening. It indicates an anomaly in the data that needs to be examined more closely.
If you did, Menzie, you’d probably come to the same conclusion as the Philly Fed…
What is left of the assumption?Well, today, the Philadelphia Fed issued a renewPutting together the official national nonfarm payrolls series from the BLS, the state-by-state aggregates from the CES, and early benchmark data from the Philadelphia Fed, we get the following chart.
figure 1: Non-Farm Payrolls, FRED Series PAYEMS (bold black), CES State Aggregate (tan), Philadelphia Fed Early Benchmark State Aggregate (red), Civilian Employment Over 16 Adjusted for NFP Concept (teal), Quarterly Employment Census and Total Wages Covering Employment, adjusted by the authors (in pink) for Census X-13, all in thousands, seasonally adjusted. Light blue shading indicates (by Mr. Steven Kopits) a hypothetical recession in the first half of 2022.Sources: PAYEMS from BLS via FRED, civilian employment adjusted for BLS’s NFP concept, QCEW from BLS, sum of state data from BLS Federal Reserve Bank of Philadelphia.
I suggest that the different years of these data confirm the following propositions: (1) the CES series track the business cycle more reliably than the aggregate employment series derived from the CPS; (2) the series are all revised over time (including QCEW).
Contact me if you see a recession in the first half of 2022. The early years of these series have been shown several times on Econbrowser; but I have yet to see Mr. Kopits admit that perhaps his assertion that the labor market data indisputably portends a recession in the first half of 2022 may be hasty.
