re: I may be missing something
You have a point that the numbers seem to need correcting for employment numbers.
The non-farm payroll numbers increased from 129,859,000 on July 1, 2003 to 131,301,000 on June 30, 2004. I run this as about a 1 percent increase.
Caveat: I can't find anything that says the BEA is using the same subset of workers that the BLS uses for calculating payroll employment. If there is overlap in the wage numbers with household employment numbers, God knows where the crossover is.
BTW, extra points for actually looking at the data. If you see any other possible holes in the blogger's analysis, let me know.