I have been following unemployment reports since 2008. I've never once thought that there was faked data in it. In fact, I'd usually be the first to admonish conservatives for accusing the data of being rigged.
There's a reason for the outrage today. What happened in today's report was nothing short of shame, and a lot of smart & nominally apolitical people recognize it.
Why? It comes down to: there was only one way to manipulate the unemployment rate with (A) it being still believable, and (B) without the market reacting negatively to it. Let's go over several ways first, to give you an idea...
The most superficial idea for tweaking the job numbers up would be just to fabricate the full time employment numbers from scratch. However, there is a fairly solid grounding of science and independent corroborations that would have pointed out the sudden creation of 850K payroll jobs to be bull. 850K job creation is consistent with a roaring economy in excess of 5% GDP growth, not with an economy with lowering consumer sentiment, slowing GDP growth, dropping manufacturing orders, and so on. So one can't really go too far by lying about the full-time jobs numbers.
Another idea is to approach the unemployment rate from the denominator part of the equation rather than numerator. This, of course, refers to labor force shrinkage, which most of us should be familiar with. While easier to pass off than just increasing the number of full-time jobs out there, labor force shrinkage is well correlated with slowing or negative GDP growth, and if the unemployment drop had been precipitated by this approach, they would have had to shrink the labor force by millions, and this would've resulted in a market panic.
So what's an Obama administration to do? They take the unscientific survey that is not taken seriously in economic evaluations, and conveniently find that many hundreds of thousands of jobs were created from part-time employment. If I were trying to engineer a big drop in unemployment while arousing minimal suspicion & churn, this is exactly how I would've done it. Very convenient.
Some websites are claiming that part-time jobs do not really get counted towards the payroll statistics, implying that you have to work a certain number of hours to get counted, hence accounting for the discrepancy between the payroll and household surveys. This is, and pardon my language, bullshit. If one is on a payroll, one is considered to have a job under the guidelines of BLS payroll survey, no matter the hours he or she worked.
The fact remains that the job report is claiming that for every one job created, eight people gained employment. There have been gaps before due to volatileness, but none nearly this much in my memory, and most of all, none so convenient, in both timing and internals of how the data was put forward.
The only honest way I can think that these statistics could have been derived is if they changed their household survey questioning to include people pawning items online as part-time merchants or something similar, but even then, a honest organization should have reported such significant changes in data gathering.
