Skip to comments.Putting Today's Job Number In "Seasonally Adjusted" Context (Read before you read the Gov't release)
Posted on 11/02/2012 6:05:47 AM PDT by Zakeet
Today's jobs number is expected to come at 125,000, with a high estimate of 154,000 from John Hancock financial, and a low of 30,000 from Westpac Banking, and with a whisper expectation at 150,000 courtesy of yesterday's "stronger than expected" ADP re-revised print. A beat or miss at over 1 standard deviation will promptly wake the HFT algos from their deep slumber. Wait did we say miss? Hah.
Anyway, just to put today's seasonally adjusted expected monthly job growth in context, below is a chart showing the average seasonal adjustment for each month of the year in the past decade. In October, seasonal adjustments subtract just over 1 million "jobs" (purely statistically of course, merely to smooth the underlying "noise").
This means that the final monthly print will be just over 10% of the actual X-12-ARIMA goalseeked statistical adjustment. Add to this another ~80K or so which will be "added" from the birth death adjustment, and one can see how in the grand scheme of things, the statistical error factor alone dwarfs what is the actual underlying data.
To think that in this labyrinth of layered adjustments to an actual number, the BLS will somehow allow the final number to be disappointing means having a locked bid on the Alaska to Russia bridge market.
Update: 171,000 Jobs Added In October, Unemployment Rate 7.9%
As expected, a whopping beat of expectations of 125,000 with 171,000 jobs added In October, and the Unemployment Rate rising modestly to 7.9%, but below the magical 8.0%. And while the U-3 rose, the U-6, or underemployment, declined from 14.7% to 14.6%. Go figure. And finally, the Birth Death adjustment came just 10K off our forecast, printing at 90K.
Simply stated, the Bureau-weenies have created another meaningless statistic designed to support liberalism using the magic of seasonal adjustments and pencil whipped job creation using the magic of birth-death models (i.e. they assume that for every company that goes out of business, a new one starts up with the same number of employees resulting in a net job loss of zero). Freepers would be well-served to assign this report the same validity as any other election eve civil servant puff piece.
The corollary number that would put all this in perspective would be to determine what the total number of people on payroll were in the previous months and compare that to the number currently on payroll.
I think that number is static at about 140,000,000 which hasn’t changed in four years and is less than the US employed just 5 years ago.
Rush is going to be bouncing off walls again today.