Nice work if you can get it . . .

The jobs report disappointed today at 92K vs. a consensus 130K. I believe this number is likely to be revised downward in the coming months because of the birth-death adjustment.

Each month the BLS has to estimate how many jobs were created in new companies and destroyed when old companies went out of business. The way it does this is by looking at how far its count was off a year ago.

That is, if last year the BLS counted 150K new jobs in July but then, after all new companies were firmly established, they went back and counted 170K new jobs they say that the birth-death adjustment needs to be 20K higher. In other words, last time they guessed 20K too few so this time they will add 20K to their baseline guess.

Here is the problem. Last time the housing market was surging and all sorts of upstart companies were coming on to cash in. Fly-by-night organizations that support builders, suppliers, etc.

Clearly those extra jobs aren't being created now and so the BLS is likely to be overly optimistic. By how much, I don't know. If there was a reliable way of fixing this problem then the BLS would use it. However, there is not and so we are left to readjust on our own.

Bottom line . . . job growth is slowing.

UPDATE: Commentors on Ritholz report that the Birth-Death adjustment for July is 26K new jobs. Typically, we find that around 50K are lost from companies' Birth/Death in July as kids get ready to go back to school, housing sales slow, etc. However, this time we got an estimated 26K created from Birth/Death. I strongly expect that to be revised downward.