Calculated Risk Sheds Light

Well I just had an interesting discussion with the guy at Calculated Risk - my mind is now completely numb.

Through the magic of the online BLS database query, new light has been shed on the mystical birth/death model adjustment, and we can all go on with our lives, secure in the knowledge that the headline jobs number is not a complete fraud.

Left unresolved however, is the question of what happens after the birth/death adjustment has worked its magic for the current month. If you look at the historical seasonally adjusted and not-seasonally adjusted data for 2004, you get 2.194 million and 2.161 million new jobs, respectively. So that kind of makes sense, the data is consistent between the seasonally adjusted and not-seasonally adjusted, it is just being smoothed - the smoother the better, as they used to say at Fannie Mae.

There is also historical data for the birth/death model adjustments, and for 2004, that total is 836,000.

So what is the relationship between the 2.1 million and 836K numbers?