Catching up on what happened as I was with my faminly Thanksgiving week and away from any computer. The weekly number of new unemployment claims was issued last week for the week ending Nobember 17, 2012, and some of the HEADLINES (MarketWatch, for example) screamed that the number was "down" ("droop") 41,0000 from the number for the week ended Nov. 10. We are still, of course, dealing with FICTION, because of Sandy, even though the jobs are really lost. However, the post-Sandy status of the job market will not be clear until, really, we get beyond Christmas and January 1. Yes, the effect of Sandy is fading, as illustrated by the "drop" of 41,000, but the holiday numbers are suspect anyway. Add Sandy, and I don't think the weeklyl number of new unempllyment claims is going to mean much until mid-January, if then. "Drops" are certainly going to mean NOTHING. IF the number shuld STAY elevated above 400,0000, that MIGHT mean trouble. 410,000 is a TERRILBE number: the highest number this year, except for th eprevius week's really distorted number of 451,000 . But the "exact" numbers each week, and the "drop" (or increase) mean ven less than usual, and the "exact" numbers NEVER mean anything. That is the main significance of the numbers announced last week: that our media and our Labor Dept. are LIARS, and incapable of giving either correct numbers or teling the truth.
Look at the week of Noveber 10. The UNREVISED number of new unemplyment claims was annunced as 439,0000. The REVISED number was announced late last week as 451,000, meaning an UPWARD revison of 12,0000 (meaning the Nov. 10 number actually INCREASED 90,000 from the preivus week, instead of the HEADLINE 78,0000). Notice the LIE in the HEADLINES thi last week, where the nuber supposedly went "down" 41,0000. It is not jsut that using "drop" or "down" in the HEADLINE is ridiculous (LIARS at MarketWatch.com did this), but that the ONLY "apples-to-apples" comparison is between the UNREVISED 439,0000 and the UNREVISED 410,0000 (to be REVISED this Thursday). Thus, IF you are gong to put this in a headline at all, you should say that the number was "down" 29,0000--NOT 41,000 . This is NOT a "matter of opinion". It is a matter of statistical HONESTY. But yoiu know by now that JOURNALISTS ARE NOT HONEST (essentialy al lof them, although I admit that--as Arthur C. Clarke said with regard to magic and sufficiently advanced science--it is pretty much impossilbe to tell the difference between sufficiently "advanced" dishonesty and sufficiently advanced sufficiently advanced STUPIDITY in "journalists") . Notice that the DISHOENST Labor Dept. has departed from its usual, CONSISTENT, "error" of 3,0000 or so to a substantially larger 12,000. "But, Skip, Sandy must make it harder.' Give me a break. The problem is NOT the "exact" number. The problem is that the REVISIN is ALWAYS distgorting the weekly number (and weekly headlines) in ONE DIRECTOIN . This is NOT a matter of randum "errors" in data, where "more complete data' changes the nuber in a RANDOM manner. The "orfe complete data" ALWAYS revises the previus week UPWARD, in the directin that means MORE UNEMPLYED. Thios means that the HEADLINES every Thursday are always BETTER than they should be: usually by that 3,0000 consistent, dishoenst error, but sometimes by a much larger number like the 12,0000 revisoni for the week ended Nov. 10. An HONEST (not to mention competent) Labor Dept. would ADJUST for the OBVIUS, CONSISTENT error in ONE DIRFECTION. I routinely adjust--doihing the job of Labor Dept. "economists" for them--by the consistent 3,000 error. I did not bother to do that in this article, because it is obviuos that the Labor Dept. unrevised numbers are even more unreliable than usual. Note how my consistent caution to you that it is a LIE to consider each week's number anything more than a fallible estimate has again been PROVEN correct. The media has LIED, week after week, in "reporting" these weekly numbers on new unemlyment claims as if they are some knd of concrete, "counting" numbers. An aboslute, indefensible, total LIE.
These weekly numbers ONLY mean anyting OVER TIME. Sandy is merely an example of the kind of "seasonal adjustment' problems, and "special factors", that occur MOST WEEKS. Again, Sandy has PORVEN me right on tis. Can you doubt that there are OTHER EVENTS that distort the numbers in any individual week? Of course you can't. Again, you MUST "report" these weekly numbers OVER TIME, or you are a LIAR (disheonst and stupid). Each week, except in context, means NOTHING. It is significant that, OVER TIME, this weekly number of new unemplyment claims has NOT IMPROVED over this ENTIRE YEAR. No. I am NOT talking abut the 451,0000 or the 410,000. You can throw those numbers out, and the equally FICTONAL 342,0000 of several weeks ago, and the weekly number this year has STUCK in the range of 351,0000-392,000. Nor has this "rangve" been distorted by "late" "improvement". The "range" was 351,0000-365,000 from mid-January to mid-March. There has been NO TREND this entire year, as we have bounced around--depending on the "season"--within this 351,000-392,000 range this entire year. Sandy will certainy cause the loss of SOME permanent jobs (although it may atually ADD some as well). But Sandy's TEmPORARY effects will make it hard to see what the longer term effects are, especially i the Christmas holiday season. Thus, we will have to see what hapens after we settle down (again, OVER TIME, and not in any one week). What yu can't do is what our DISHOENST media is waiting to do: procalim an 'improvement" as the numbers drp back below 4000,000 (assuming they do). Again, the job losses from Sandy are REAL job losses, as distinguished from the ERROR by the DISHONEST Labor Dept. when it failed to count thousands of claims in California one week (resulting in that FICTIONAL HEADLINE of a "4-year low" of 342,0000). The job losses from Sandy are REAL. They just MAY be mostly TEMPORARY. But our DISHOENST media refuses to look at "special factors" unless the distortion is BOTH obvius, AND in the "wrong" direction (as with Sandy).
Progress? Not much, but this blog MAY be having some effect with my (correct) RIDICULE of our media and our Labor Dept. One CNN article (although probably not the truly DISHOENST peole on TV) actually erferenced the number of new unemplyment claims for the week ending Nov. 17 as "about 410,000", while the headline emphasized the effect of Sandy . Even Marketwatch,.com, with the truly disgraceful headline of LIARS, used the word "estimate" in its description of the annunced weekly numbe of new unemplyment clamis. That should be IN THE HEADLINE of EVERY story on the weeky number of new unemplyment claims: that it is an ESTIMATE. And the body of EVERY article should make clear that the weeky numbers mean NOTHING except OVER TIME. No. If you "media peole" think peole in general are TOO STUPID to handle actual attempts to give the "fuzzy" facts, rather than "black and white" LIES, then you are DISHONEST as Hell. What am I saying? You ARE as dishonest as Hell.
Oh. Then there are "economists"--supposedly "surveyed" by Marketwatch each week to PREDICT the weeky number. Now, the "predictions" are NEVER right, UNLESS the number hardly changes. But do you really want to know how DUMB our "economists" really are? Those "economists" surveyed by Marketwatch actually MADE A PREDICTION for the week ended Nov. 17, 2012. Now any HONEST "economist" would have to rEFUSE to make a "predictin" for that week. Ridiculous. Sandy obviusly is having UNPREDICTABLE (except perhaps in DIRECTION) effects on the weeky numbe of new unemplyment claims. What kind of IDIOT would make a PREDICTION in this situatin? Right. A DISHOENST IDIOT. Yep. I jsut said that Marketwatch exposed its "economists" "surveyed" as both DISHOENST and STUPID. But they had already done that week after week .So it is not really anything new, and something I have told you YARS ago.
P.S. No proofreading or spell checkng (bad eyesight).