I was sincerely hoping for good employment news today, to confirm all the good economic data points we've seen in the past two weeks. I've never believed the weekly first time unemployment claims data, although the one-week later revisions might be a bit better. But the numbers never seemed "quite right" to me and recently my doubts were confirmed when the
AP and others reported that yes, indeed, the numbers are massaged and manipulated, on occasion.
Even if the numbers are bogus, they are data points that can be followed and trends spotted (sort of like IPs in the Bakken).
Last week, the unemployment claims number was almost too good to believe, but with the other good academic data, the numbers were plausible.
I was eagerly awaiting today's figures.
I guess the analysts forgot that every year about this time there is a national holiday. What amazes me is that one holiday -- one day off from work -- can affect unemployment claims to this extent. Are folks that quick, that as soon as they lose a day's work due to a holiday -- they are in the unemployment office submitting a claim for benefits? I guess so.
Look at these numbers. I have never seen a surge to this extent in unemployment claims in the two or three years that I have been
tracking these numbers. Scroll through those three pages and see if one can find a surge of 68,000 claims in one week due to a single-day holiday.
Generally the number moves 2,000 to 5,000 up or down. Very rarely, do we see moves of 20,000 up or down. I do not ever recall a number this high "blamed on" a scheduled event.
That was the headline: "surges 68,000."
Let's see how
Bloomberg, a mouthpiece for the administration, handled this shocking news:
Applications for U.S. unemployment benefits jumped last week from an almost three-month low, highlighting the difficulty in adjusting the data around the year-end holidays.
Jobless claims surged by 68,000 to a two-month high of 368,000 in the period ended Dec. 7, exceeding the highest forecast in a Bloomberg survey of economists, Labor Department data showed today in Washington.
The 300,000 applications filed in the prior week, which included Thanksgiving, were the fewest since Sept. 7.
The data reflect seasonal adjustment volatility around the Thanksgiving and Christmas holidays, a Labor Department spokesman said as the figures were released. A report last week showed the unemployment rate fell to a five-year low and companies added more workers than forecast, pointing to further labor-market progress.
Ah, yes. This is the
Bloomberg spin: Forget the actual numbers. Don't spend any time analyzing what this means for the economy. Rather, analyze why the analysts got it wrong. It was due to difficulty of forecasting job numbers around scheduled holidays.
Really? They've been doing this for decades. The number (68,000) was higher than the highest forecast. Missing by this much is akin to Asiana pilots landing short of the San Francisco runway.
If folks recall: stores opened earlier than usual for Black Friday, opening up on Thanksgiving. One would assume that more workers were hired, not fewer.
Nah, something is going on, but it's not a single day holiday.
Also, note this in the
Bloomberg, an administration mouthpiece: at the very top of the story they reminded readers how good the numbers were just a week ago, suggesting this was a one-off OR that things were really not this bad. A suggestion, subliminal suggestion, to ignore this data.
This is sort of the clincher:
“I wouldn’t put too much stock in the ups and downs of initial jobless
claims over the next several weeks because seasonal volatility is pretty
high this time of year,” said Ryan Sweet, senior economist at Moody’s
Analytics Inc. in West Chester, Pennsylvania,
and the top-ranked forecaster of jobless claims in the past two years,
according to data compiled by Bloomberg. “Layoffs are low. Other jobs
data suggest layoffs are not the problem, it’s the lack of hiring.”
Really? Just yesterday (or the day before or the day before that) the mainstream media was telling us how good hiring was getting.
I can't make this stuff up.