More updates on the Birth Death Model:
Birth Death Adjustment Goosed NFP
Payrolls revised, mistakenly said unemployment rate fell | The Big Picture
What continues to amaze me is that here we have better than expected job loss data, but somehow, the birth/death (B/D) model generated 175K more jobs in May 2009 compared to May 2008. I am always, always, always skeptical of "facts" obtained from simulations and models. Why? I used to do a lot of molecular orbital work. The results were never very good on the whole, although we achieved some successes on a hit or miss basis. Based on my own work, and the work of weathermen predicting rain in my front yard, I am even more skeptical of something called a birth-death model of jobs creation. (Sounds like it's releated to an atmospheric CO2 model for the year 2020 with "seasonal adjustments" thrown in.) The government says it's just a statistical fluke. Here's the government's explanation of the Birth/Death model: CES Net Birth/Death Model
This link helps explain some of the government modeling in more detail:
Impact of business births and deaths in the payroll survey (PDF)
Not sure about this fellow, but he offers some more detailed analysis. I visited the BLS web site, and all I can say is, gerbils must have come up with the model. The results and historical data for the B/D model is all over the map. Good luck believing the results.
May Employment Report Not Believable
Benjamin Disraeli, where are you when we need you?