For those as obsessed with such things as myself, you know that before we get the big jobs number on the first Friday of the month, the private payroll processing firm ADP releases their own estimate of last month’s private sector job growth.
Their estimate for August is a higher-than-expected 201,000 jobs. That’s an acceleration over their July number of 173K, revised up from 163K.
BTW, if that 163K sounds familiar to you it’s because that’s the same payroll number from the BLS report last month, though that number includes the public sector too (the comparable private sector number from the BLS payroll survey for July was 172K).
So, the question is: can you jump from the ADP to the BLS number on a monthly basis?
If you look at the correlation of the monthly changes in the series—0.95—you’d think so, but alas, while they move closely together (ergo the strong correlation), in any given month they can be quite different and the differences can go either way.
For January of this year, the ADP reported 170K jobs. A couple of days later, the BLS came in at 257K. A couple of months later, in March, ADP came in at 209K and BLS followed with 121K. (These are all first-prints, pre-revisions.)
The figure below shows the ratio of the first prints of the ADP over the BLS, monthly changes in private sector employment (so ’1′ would mean they were the same). As you can see, they jump around from <1 to a lot >1.
So all you say is the consensus for tomorrow is around 130K on payrolls, with, thanks to the ADP pop, an upside risk. Me and Jobbsie (that’s my jobs day monkey, whom I blindfold and spin around a few times before he throws a dart at the number line) are coming in at 150K.
That number is consistent with real GDP growth that’s around trend and even perhaps a touch optimistic given the recent acceleration in productivity growth. It’s also a good number but not a great one. Like I said, we’re definitely adding jobs—especially, conventioneers, compared to four years ago!!—but we need to be adding more.