THEME: UNEMPLOYMENT
Interesting comment made by Alan Abelson of Barron's...

Last week the Bureau of Labor Statistics dutifully came out with its February jobs report. And once again, the consensus -- which in this case is heavily weighted with con over sense -- woefully overestimated the additions to payrolls. The going guess was anywhere from 130,000 to 200,000, and there were plenty of mutterings about the risk in the estimate being very much to the upside. In fact, a stingy total of 21,000 new jobs were created in February, and except for hiring by kindly government, not a single one would have been added at all.

OK, so we don't expect the errant Nostradamuses to indulge in a mass frenzy of self-flagellation at the corner of Broad and Wall. But maybe a tiny hint of fallibility? Not on your life. If their figures are wrong, then there must be something wrong with the figures. Why, of course.

In particular, the seers insist there's nothing amiss with their prophecy; rather, the rest of the world is looking at the wrong data. It isn't the establishment numbers, stupid, they scoff, but the household numbers that count. Need we say, the household numbers have been running substantially higher than the establishment numbers. (As you've doubtless surmised, the Labor Department does two surveys of unemployment, one a sample of 60,000 households conducted by the Census Bureau, the other a sample of 160,000 businesses and government agencies, collected from payroll records.)

For some reason, release of the February jobs report did not occasion the usual refrain. Perhaps it had to do with the fact that according to the household survey, the nation lost 265,000 jobs last month. We might add that anyone who makes even a glancing examination of the two surveys -- and, some years back, Barron's did a wonderful reportorial job on the nitty-gritty of how the data were collected for each -- is inevitably struck by how shaky, even flaky, the household polling can be.

What the sunshine crew can't blink is that on virtually every score this was a sorry jobs report. Making it all the glummer was that the payroll gains originally reported for both December and January both were shaved significantly. January's 112,000 rise was trimmed to 97,000; December's already feeble 16,000 was sliced in half. In other words, the job picture not only was gloomy last month, but also was dimmer than thought in the prior two months.

Manufacturing extended its long losing streak in February, dropping another 3,000 slots. Jobs in this beleaguered sector have now slid 43 months in a row, and counting.

The only reason the unemployment rate stayed at 5.6% was that the labor force shrank by a hefty 392,000 souls. As Wells Fargo's Sung Won Sohn observed -- after sighing that last month's paltry payroll additions were "a terrible number" -- instead of such pitiable increases, almost three years into recovery the economy should be churning out 200,000 to 300,000 jobs each and every month.

Beyond the headline numbers, the picture is even darker. The average stretch that workers losing jobs stayed idle hit a new 20-year peak in February. If you toss in the folks who quit looking for jobs because they've been searching fruitlessly and are thoroughly discouraged, plus those who are eager for work and have tried recently but failed to land a job, along with people who are working part time because they can't find full-time occupation, the unemployment rate rises to a formidable 9.6%.

Arguably, that's a truer measure of just how jobless this recovery really is.