The New Unemployment Numbers: Green Shoots?

Are those "Green Shoots" finally poking up through the snow?

According to the Bureau of Labor Statistics, we may have finally
turned the corner on Unemployment.

In its monthly statistical report released Friday, the
number of "jobs lost" in November was a mere 11,000,
while the overall unemployment rate declined from
10.2% to 10.0%.

Judging by the response, the mainstream media and
the financial Powers That Be were falling all over themselves
in self-congratulation that the "Worst is Behind Us".

The irrepressible (and in my mind completely irresponsible)
Jim Cramer was celebrating the end of the "Jobless Recovery",
along with the usual self-serving promoters and shills in the
chorus.

The Administration also took this as signs that
"its policies are working" and started thumping the tubs
for its "Employment Summit" later this week.

To all of those with a stake in the status quo, things
couldn't be looking better. And those with an immediate
stake in some economic good news, like embattled Fed
Chairman Ben Bernanke, started grasping at this
straw with the hope of a drowning man clinging to driftwood.

But the question does remain - have we turned a corner?
Or is this merely a statistical aberration? Or, as the conspiracy
theorists have it, have these numbers been doctored to achieve
a certain outcome?

My take on the above: No, Yes, and Maybe.

On the first question, "have we at last turned the corner?"
my vote is No.

To begin with, this remarkable decline in the rate of job
loss is unconfirmed by any other statistical measure.
Hours worked for both manufacturing and non-manufacturing
employees is up from 33.0 hours per week to an anemic 33.1 ,
well below full employment numbers and well within
the range of statistical error.

Other measures of economic activity - rail and trucking carloads,
inventories, and new orders - are flat to slightly down.
Thus, this remarkable piece of "good news", while welcome,
has to stand as unconfirmed by any other measure.

In fact, if you want to take a very pessimistic stance and
drill down a little bit, the complementary statistics for
unemployment are as dim as ever. The percentage of
"counted" unemployed that have been unemployed longer
than 36 weeks (9 months) is 38.6% of all unemployed - the
highest number in 25 years. And while the number of those
unemployed for less then 9 weeks has dropped
significantly (from 15.3% of total unemployed to 11.6%)
over the last three months, the other Quartile measurements
of unemployment duration have been growing.

What does this mean? It means that people, once unemployed,
are moving to the ranks of the long-term unemployed - and
staying there. This is also also confirmed by the broad measure
of unemployment U6, which counts the unemployed,
the discouraged, and the crucially important "unemployment
effect of reduced hours" (i.e. for every four people cut back from
40 hours to 32, add one to the "currently unemployed" head count).

That number remains a depressing 17.2%.

Which brings up the second question - is this miraculous
Unemployment figure a statistical aberration?
I think we can safely say Yes.

To begin with, any raw count of U3 (the official
unemployment rate), has a Mean Statistical Error
of +/- 100,000. This number has remained very consistent,
even adjusting for its dependent variables labor force size
and labor force participation rate. So, allowing for error,
the true number could be anywhere from -11,000 to -111,000,
depending upon sample size and methodology, both of which
the BLS is free to alter at any time.

Further, when you take into account Seasonal adjustment,
which normally adjusts LFS (labor force size) and LFPR
(labor force participation rate) upward to account for seasonal
Holiday retail and shipping employment, this purportedly good
report is skewed further.

So, to confirm whether or not this "good news " is real,
I'm going to look at a proxy for the BLS numbers - The ADP
payrolls report, which measures an approximation of U3 through
payroll rather than household data.

The ADP numbers for November? Job Loss = -169,800
(still dismal, but improving), with a Holiday Seasonal
adjustment of +52,000.

Conclusion? Statistical and seasonal aberration most likely,
with all numbers within allowed statistical error.

Finally, we come to the "Conspiracy Theorist" outlook,
most frequently bandied about by websites such as
Zero Hedge, Naked Capitalism, Calculated Risk, and others.

Basically, the "conspiracy" meme runs like this - The
Administration, Congress, The Fed, and the TBTF banks
(most prominently Goldman Sachs), were going to drastically
cook the November unemployment numbers to mislead the people.
And on the surface, there's some justification for this.

The forecasters were predicting that retailers were facing
the worst Holiday retail season in a generation. More retail
chain bankruptcies were predicted for the first quarter of 2010,
with all the attendant negative fallout for employment and
commercial real estate.

That's not what the Powers That Be want to hear, goes
the theory, so watch for an "October Surprise" - namely,
unemployment numbers that would be miraculously
optimistic, in order to goose consumer spending.

Did that happen? In my opinion - Maybe, but only because
this relatively meaningless statistical improvement has
been flogged beyond all credibility as being "positive" by
mainstream media types who should know better.

Even Goldman Sachs - the hated Squid - put out its
November expectations as an improving
(but still negative) -101,000, with a qualifier to not
expect much improvement in the first or second
quarters of 2010.

This, in my mind, is consistent with the other data.

So - in conclusion? Things are still bad, but slowly getting
less so, and the economy will continue to bounce along
the bottom for the next several months.

Now that is not good news, for the politicians especially - but
we'll cover that later. And as for the "Green Shoots?"

Relax, they're not here yet.