Okay, so the December US jobs report is out, and it's much worse than expected, with
85,000 jobs lost counting seasonal adjustment. (
Canada lost jobs as well, similarly expecting a gain, but the situation wasn't as bad. Thanks to
cow for the link.) Most people were expecting a net gain. Note that to keep up with population growth, you need to have about +100,000/month to tread water w.r.t. the headline unemployment rate.
Why did the unemployment headline number (U3) stay at 10.0%? Because a lot of people left the workforce entirely, and it ended up about a wash. Technically the U3 got very slightly worse, but it still rounds to 10.0%. From the BLS report:
The civilian labor force participation rate fell to 64.6 percent in December. The employment-population ratio declined to 58.2 percent.
The U6 report - total un- and under-employment, including discouraged workers - ticked up 0.1 to 17.3%, seasonally adjusted; the
non-adjusted number jumped 0.7 to 17.1%.
The December report of 85,000 jobs lost includes 59,000 non-seasonally-adjusted BLS birth/death model jobs added. For those coming in late, the Bureau of Labour Statistics's birth/death model attempts to estimate jobs gained and lost through to business startup and failure. It is (or at least, can be) unique to every month, so month-to-month comparisons are essentially impossible.
It has only gone negative once throughout this entire recession, and it is a sham.
Looking back to last month, November was revised up from a small jobs loss to a small jobs gain (+1000, seasonally adjusted). However, as you'll recall, the completely broken "birth/death model" added 30,000 jobs to the non-adjusted tally that month. No matter what adjustments the BLS did for seasonality, take that back out, and November's still going to be a negative number. (Here's
Mish's monthly roundup on the numbers, with all his convenient tables.)
And if you think it's just cranky people like me and Mish harping in birth/death, it's not; I missed this back in October(!), but
the BLS has admitted the model is completely broken, and estimated at that time that it had overestimated job growth in this recession by 824,000 as of March 2009. That's more than all the jobs added by the birth/death model throughout January through October 2009. It's all going to be taken back out in the January numbers, which come out next month. Plus all the months after that have been positive as well; assume those are generally negative as well.
So that's going to be quite the crater. One might say it'll leave a
mark. Mish has
an assortment of commentary here. This won't affect the headline rate, though; it's a different survey, and there are legitimate reasons for that. But it will affected the jobs-lost numbers.
More frustrating is that
those without jobs continue not to find new ones, in large numbers. The only reason there isn't rioting in the streets from joblessness, frankly, is the Extended Benefits and Emergency Unemployment Compensation benefit extensions. Seriously,
here's the graph showing "continuing claims" - standard benefits. Note that "continuing claims" have fallen. (Also note that graph isn't zero based so it hasn't fallen as much as it looks like it has.) Looks good, right? Well,
add that to this chart of Extended Claims benefits, filling in most of the gap until the last few weeks. But now that's falling - so that's good, right?
Not when you see that the Emergency Claims rate has spiked to fill in that gap, as well. (All charts courtesy Chris Puplava at Financial Sense, via Mish Shedlock, thanks to both.)
And all this helps explain why
December withheld tax receipts fell yet again, rather than rising, as expected, and why
state tax revenue through September 2009 dropped the most since 1963.
November's job report genuinely wasn't too bad. December's was not attractive at all. January's is going to be a mess due to BLS revisions, so it's going to be fun times making heads or tails out of
that.