From the Department of Labor, announcing a 5-year low in adjusted jobless claims:

“The advance number of actual initial claims under state programs, unadjusted, totaled 436,766 in the week ending January 19, a decrease of 119,944 from the previous week. There were 416,880 initial claims in the comparable week in 2012.” [emphasis added]

According to the statisticians at the DOL, jobless claims actually rise Y/Y, but on a seasonally adjusted basis it’s a five year low? That’s a non-trivial 4.8% jump. For the record, the seasonally adjusted number for this week in January 2012 was 372,000.

So, according to the DOL’s adjustment math:

416,880 actual claims in 2012 = 372,000 adjusted
436,766 actual claims in 2013 = 330,000 adjusted

FWIW, I think the fact that Gallup is now actively highlighting and advertising better accuracy in their employment survey vis a vis the DOL due to absence of adjustment processes and estimates indicates not just the growing incongruity of data manipulation (bad theory, not outright lying) in these statistical adjustments, but, more importantly, heightened awareness of this nonsensical fudging. Perhaps some leeway is afforded in that January is often a difficult month for any estimate set, but the trend-cycle analysis is obviously showing where it should be beyond inconspicuous.

Statistical fudging aside, the most important data point in the release is likely the 214,000 that fell off UI completely (just this week). There are 2 million fewer claimants this year than last, plus 2 million MORE added to the civilian non-institutional population. Even according to the DOL’s Establishment survey, there were only 2 million total jobs added last year (and many were part-time). That leaves 2 million potential workers (plus their families) without any source of institutional or earned income. SS Disability absorbed a few ten thousand, but that still leaves several millions (again, plus their dependents) with nothing.

No need for an exogenous shock to fall into recession.