Possibly good, the ever-optimistic Ed Yardeni of Yardeni Research says this morning:

I nearly fell off my chair yesterday morning at 8:30 am when I heard that the latest weekly initial unemployment claims fell below 300,000 for the first time since May. Actually, I was driving to meet with a couple of accounts in New Jersey and heard the news on Bloomberg radio. I don’t get it. Isn’t the labor market supposed to be getting weaker? That’s been the message of every labor market indicator recently with the glaring exception of jobless claims. Perhaps employers aren’t hiring, but they are resisting firing workers because they believe their businesses will be good next year. Or maybe today’s job losers--especially the ones who worked for a subprime mortgage unit on Wall Street--are too proud or too well off to bother filing for unemployment insurance. In any case, the one good thing about August’s 4,000 drop in payroll employment is that any increase above zero would likely be a big positive surprise. Just imagine the reaction if September’s number, which will be released on October 5, is up by let’s say just 50,000. The newspaper headlines would probably hail that the labor market remains surprisingly resilient. There might be dancing in the streets--both Main and Wall--because the economy may not be heading toward a recession, notwithstanding the predictions of Alan Greenspan, Richard Syron (Freddie Mac's CEO), and other gloomy oracles.

Not as good as you think, Merrill Lynch's sober David Rosenberg said yesterday:

Falling jobless claims does not equal labor market strength. No doubt that initial jobless claims surprised to the low side today - dropping 15,000 to 298,000 in the September 22nd week while the consensus was looking for a 316,000 print. This latest data point also dragged the 4-week moving average down to 312k from 321k and the lowest since 4 August 2007. This in turn has prompted comments in some circles that the labor market is actually stronger than the recent nonfarm payroll survey indicated (-4,000 in August). We disagree.

Aggressive seasonal factors yielded sharp drop in claims First, as our bond desk rates strategist John Herrmann pointed out earlier, the previous four years of extremely heavy hurricane activity caused the seasonal adjustment factors for initial claims to be "higher" than they would otherwise be during a period of benign weather patterns as is the case this year. So the seasonal factors were looking for a bump-up in the raw claims data (i.e. the seasonal factors would have attempted to lean against what was supposed to be a sharp rise in the seasonally unadjusted numbers), but that weather-induced runup didn't happen this year. So much of the surprise drop in claims owed to aggressive seasonal factors, not to any incipient strength in the labor market.

Initial jobless claims tell you only about layoffs Even if the claims data had done what the consensus had expected, they still would have been quite low at less than 320,000. In fact, even as job market conditions have slowed in the past year, the claims data have been quite benign. This then begs the question - why are claims so low? Well, initial jobless claims only tell you about layoffs - these are the folks heading to the unemployment insurance line.

Employment changes are about hirings and firings Employment changes are determined not only by the number of firings in any given month but also by the number of new hires. And the jobless claims data do not capture new hirings - and that is why there has been a disconnect between this data series and the official employment surveys. The firing rate is indeed low, but the hiring rate is also is low (in other words, labor market turnover has receded sharply in the past year) - and the latter is actually slowing down faster than the former which is why the NET job creation figures are so anemic.