I am AMAZED at the innumeracy running rampant in America.
ABC runs a story that says, in part: “U.S. employers dramatically stepped up their hiring in December, adding 312,000 jobs in an encouraging display of strength for an economy in the midst of a trade war, slowing global growth and a partial shutdown of the federal government.” And that is one of the leads on Drudge.
Oh, and the market screams up 600 points.
Fine. Except, we don’t find that number in top line (Summary Table A) of the US Department of Labor Report (here). have a look:
Yes – there’s the DoL data showing that for the month the number of working was up 142,000.
The way ABC got the headline-data was by going to the (noisier) Establishment data in the B Table which is not nearly as meaningful as the Summary. Note my three highlights:
The first highlight was the apparent basis of “312,000” jobs headlines.
Worse? My second highlights show zeroes in November for mining, logging, and construction..
Importantly, Establishment data doesn’t tie-back neatly to the Summary. Noisy data series.
Elsewhere, the grand market-spurring hype is that “Hispanic unemployment is the lowest ever.” Well, sort of. But that’s problematic in and of itself…
My highlights here show the Hispanic gain in “Employed” was up 177,000 in December.
Oh-oh. See where this is going? In the Ure view of Reality (backed-up by data) we see a case that is NOT in the MainStream Media – which is clearly (to us, anyway) cheer-leading to save the Stock Market.
Since TOTAL Employment in America really was up 142,000 persons in December and the unemployment rate rose. That doesn’t bother anyone, I ‘suppose.
More to the point, Hispanic employment was up 177,000 persons…want to tell me what this spells for the rest of us?
Our headline might have been “Non-Hispanic Employment fell in December by 35,000 people.” And that, dear reader does tie-back to unemployment going up.
Perhaps the MSM should stick with “standup and pancake” or go back to school.
Anyone up for a refresher on inferential statistics? Or, how about “How to read Descriptive Statistics?