And throughout much of 2010 and 2011, BLS frequently discovered more positions, and revised the jobs figures up.
Each month when BLS reports results, it looks backs and revises the prior two months.
The household survey, which is the other survey the BLS does but gets far, far less attention, was weaker still.
Finally, we looked at the average weekly hours worked in each city, also from the BLS.
And then, each year, BLS goes back and revises all the data for the prior 12 months.
Put another way, the BLS looked back and found there were 74,000 fewer jobs than previously thought.
BLS polls a sample of firms and then extrapolates those figures to reflect the experience of all firms.
For example, this month, BLS revised the October jobs gain upwards, from 203,000 to 241,000.
The talk-radio right will start to lay into the BLS and try to discredit it.
The BLS report, like most pieces of economic data, is an estimate.