The ADP report gives us a first glimpse at what employment did in the month of June. The official payroll series comes, as always, on Friday.
The best we can say is that the series is moving in the right direction. There are fewer job losses this month than the last. That is an ambiguously good sign. The problem is that the rate of improvement is slowing. Yes, that’s right I am bringing up the third derivative, changes in the rate of change of the change in total employment. These things matter, however, because the derivatives not only help us forecast future values but display patterns themselves.
Even in the last jobless recovery we can see that the change in employment rockets up close to zero before slowing down. That is, the economy moves quickly towards neither creating nor destroying jobs. This time the economy is giving hints that it wants to stay out around minus 400K a month. As if the steady-state where a recession.
With the specter of the Japanese Lost Decade still looming, such signs are less than encouraging. Still two months don’t make a trend and a month’s worth of third derivative analysis makes even less than that. Stay tuned.

2 comments
Comments feed for this article
Monday ~ July 6th, 2009 at 5:24 pm
Tim Fowler
“Changes in the rate of change of the change in total employment”
Wouldn’t that be the 2nd derivative? With the rate of change being the first, and total employment not being a derivative at all?
Monday ~ July 6th, 2009 at 5:58 pm
Karl Smith
Tom,
I think it goes this way:
Changes in the rate of change are the second derivative on the graph. That is the slope is decreasing and thus the line is concave down.
However, the graph itself is of the first derivative of employment.
Thus we have the third derivative of employment.
Or in other words. Employment is declining (1st) but at slower rate (2nd) but the rate of slowing is itself slowing (3rd).