Mr. Steven Kopits asserted that the Philadelphia Fed’s early preliminary benchmark supported a recession in 2022H1, to wit:

You, Menzie, held the Est Survey was more likely right. You wrote: So: (1) I put more weight on the establishment series, and (2) the gap between the two series is more likely due to increasing, and biased, measurement error in the household series, rather than, for instance, primarily increases in multiple-job holders.  https://econbrowser.com/archives/2022/12/the-household-establishment-job-creation-conundrum

Dead wrong, as it turned. And predictably so.

You were wrong because you did not consider the statistics more holistically. That’s the learning point for your students. Cross check your indicators if you have dials which are telling you different things. If jobs are increasingly rapidly, then GDP should also be up. If jobs are increasing rapidly, then mobility and gasoline consumption should also be up, because so many people need to drive to work in this country. Finally, if productivity is imploding when jobs are up, you really need to take a pause and put together some sort of narrative as to why that might be happening. It suggests something anomalous in the data which requires closer inspection.

Had you done that, Menzie, you might have concluded as did the Philly Fed…

What remains of that hypothesis? Well, on March 16th, the Philly Fed released this update.

Continue reading →