Wednesday, February 2, 2011
I have always hated the fact that the monthly jobs number ends up being so important. Whatever the number is that is announced this Friday by the BLS is almost sure to be revised significantly, up or down, in the months and years to come, as BLS eventually reconciles its estimates to actual data based on IRS tax receipts. Sometimes the revisions can be of an order of magnitude in size.
It's also the case that the ADP monthly estimate of the number of private sector jobs gained or loss has lost a lot of credibility in the past year or so, since it has ended up grossly underestimating the number of jobs that the BLS has reported. You can see this in the chart above, where the red line has been consistently higher than the blue line for most of the past year.
So are both these numbers so flaky that it is not worth paying attention to them? Only time will tell, but in any event you have to take them with several grains of salt at least. Both can be wrong.
Nevertheless, it is tempting to get a little excited about the relatively large January jobs gain that is being projected by ADP in their release today (+187K), since estimates for the BLS number are only +140K, and 200K jobs per month is an important threshold number since that is what is required—if it persists for a sufficient period—to bring down the unemployment rate. ADP overestimated job gains last month by a lot, but that was the only time they had overestimated the BLS number in over a year. What if ADP once again is underestimating the BLS number? If they are, then Friday's payroll number could be a welcome surprise to the upside.
I wouldn't want to put much stock in these speculations, however. As anyone knows who has followed at this data over the years, trying to predict monthly jobs numbers is fraught with difficulties, and the value of the exercise, even if one is right, can be negated months later by revisions to the original data. I only mention this because it is tempting, and because there are other indicators which are pointing to a substantial pickup in jobs numbers going forward (e.g., declining unemployment claims, stronger-than-expected corporate profits, stronger leading indicators, strong growth in capex, strong growth in commodity prices, strong export growth, strong car sales).
Posted by Scott Grannis at 3:16 PM