The jobs report seems to bring out many myths which I'd rather just debunk with some facts before we get there.
Myth: There is an extraordinary gap btw payrolls and HH survey growth rates
Fact: adjusting for definitions suggest is a tad high of late but not abnormal
Myth: There is an extraordinary gap btw payrolls and HH survey growth rates
Fact: adjusting for definitions suggest is a tad high of late but not abnormal
Myth: the HH survey is better at predicting turns than payrolls, particularly the first report.
Fact: The divergence between the two measures (adjusted for definition differences) is random and has no leading signal one way or the other through time.
Fact: The divergence between the two measures (adjusted for definition differences) is random and has no leading signal one way or the other through time.
Myth: An extraordinary number of people are desperately taking second jobs. That is a negative and explains the divergences we see.
Fact: Second job holders has been ~7mln for 2 decades, rose 500k in the last 6m, and if anything second job taking is a positive cyclical sign.
Fact: Second job holders has been ~7mln for 2 decades, rose 500k in the last 6m, and if anything second job taking is a positive cyclical sign.
Myth: Self employed jobs are collapsing which accounts for the difference.
Fact: Self employed is pretty stable at 9mln and choppy. It did fall in the last 6m explaining about 500k of the diff, but wouldn't read much into it given the vol in the measure.
Fact: Self employed is pretty stable at 9mln and choppy. It did fall in the last 6m explaining about 500k of the diff, but wouldn't read much into it given the vol in the measure.
Myth: Payrolls get revised massively in the opposite direction in subsequent months so you cant trust first release.
Fact: Payrolls are typically revised to go more in the direction of the original release, particularly in extreme events. Recently its been roughly zero.
Fact: Payrolls are typically revised to go more in the direction of the original release, particularly in extreme events. Recently its been roughly zero.
Myth: Payrolls get revised down a ton eventually so they are meaningless to understand now.
Fact: Payroll revisions are pretty modest and typically in the direction of the original payrolls.
Fact: Payroll revisions are pretty modest and typically in the direction of the original payrolls.
Myth: Payrolls overstated job gains by 1mln in 2Q22.
Fact: The approach by the Philly Fed has been widely discredited because it has issues with seasonality and implied an implausibly large adjustment relative to history.
Fact: The approach by the Philly Fed has been widely discredited because it has issues with seasonality and implied an implausibly large adjustment relative to history.
The best path isn't to try to be cute. Look at both, adjust for their confidence intervals (HH survey is wider) and get a probabilistic sense of reality, which works out to in between the two releases, a little tilted to payrolls:
Loading suggestions...