Some less-than-stellar news on the unemployment front. The Unemployment Insurance Weekly Claims Report has been released. Initial claims grew to 473,000 while last week’s figure was revised up 2,000. This was above the Bloomberg concensus range of 410k to 450k. Just to provide some positive spin: the 4-week moving average still fell, though not to where it was the second week of January (440k, the best point since August 2008). From the report:

In the week ending Feb. 13, the advance figure for seasonally adjusted initial claims was 473,000, an increase of 31,000 from the previous week’s revised figure of 442,000. The 4-week moving average was 467,500, a decrease of 1,500 from the previous week’s revised average of 469,000.

The advance seasonally adjusted insured unemployment rate was 3.5 percent for the week ending Feb. 6, unchanged from the prior week’s unrevised rate of 3.5 percent.

The advance number for seasonally adjusted insured unemployment during the week ending Feb. 6 was 4,563,000, unchanged from the preceding week’s revised level of 4,563,000. The 4-week moving average was 4,585,750, a decrease of 24,000 from the preceding week’s revised average of 4,609,750.

I’m sticking with my story: New claims must fall below 400k. In the last recession, new claims only rose as high as 517k. They recovered a bit, but hovered around the 400k point. The recovery in jobs didn’t begin until the recession was well over, after the new claims number managed to move convincingly below 400k. Census hiring may provide some temporary relief, but don’t expect permanent, structural improvements until we see employment volatility come down. I’ll throw in the painful reminder that current estimates are that we need 125,000 new jobs a month just to keep up with the population.

The unadjusted numbers, those that correspond to actual people instead of trend data, give us better news. The seasonal adjustment caused by the layoff spike at the beginning of the year (every year) is officially over. The adjusted/unadjusted spread is only 3,000:

The advance number of actual initial claims under state programs, unadjusted, totaled 476,730 in the week ending Feb. 13, a decrease of 30,850 from the previous week. There were 619,951 initial claims in the comparable week in 2009.

The advance unadjusted insured unemployment rate was 4.3 percent during the week ending Feb. 6, a decrease of 0.1 percentage point from the prior week. The advance unadjusted number for persons claiming UI benefits in state programs totaled 5,539,706, a decrease of 150,689 from the preceding week. A year earlier, the rate was 4.5 percent and the volume was 5,972,146.

I’ve plotted the seasonal adjustment for the last 10 years. Jobs that are added are later taken away, with two spikes of jobs added, and two spread out periods of jobs taken away. The adjustment graph for the monthly employment report looks similar. And this is why the Census hiring is a big unknown. It appears as if the bulk of the hiring will come in during a period when the seasonal adjustments take away jobs. As a side note, it’s far easier to present these complicating factors than it is to try to digest them and make a reliable prediction that accounts for them. Hopefully it is also more meaningful, as knowing the factors that go into the summary numbers provides a more complete picture than the summary numbers themselves, which tend to have a magical random number quality to them.

The good / bad lists (corresponding to the first week of February) show decreased volatility.

The good list (-1000 or more): CA, PA, FL, IL, WI, NC, TN

The bad list (+1000 or more): OK, AR, MI, IA

IA (the worst) was +2,014 vs CA (the best) at -13,535. The big 4, construction, service, manufacturing, and trade, were well represented in the good list. Not many conclusions can be drawn from the bad list.

In other news, stories are starting to circulate about the impending extended benefits expiration. This is still in the low-panic mode, as the numbers cited are only in the hundreds of thousands (in reality, the problem will still effect around 3 million), spread over a duration of months (phrases like “by June” are being used). If the date is correct, expect the frequency (and accuracy) of the stories to increase over the next three months.

The short term outlook is probably dominated by the expiration of COBRA subsidies on the bad side and census hiring on the good side. The long term outlook continues to be dependent upon the only-obvious-in-hindsight emergence of an engine for jobs growth.