The markets got a bit of a boost with a surprise announcement that initial jobless claims dropped to a seasonally-adjusted 398,000 for the week ending July 23, down from an upwardly-adjusted (by 4,000) 422,000 the prior week. Because once again, the “experts” weren’t even close, Reuters broke out a variant of its usual adjective to describe this. This time, however, it’s because the “experts” missed rather badly on the high side – they expected the number to come in at 415,000.

The first question is whether it actually is under 400K. That’s actually a two-part question. I will say that a significant part of the drop is real – the non-seasonally-adjusted number was a bit under 367K last week, a drop of 103K from the prior week. I’ll leave it to the pros and the tin-foil-hat types to argue whether 398K is a “real” number or a political one that “conveniently” popped up when it did.

Is it going to end up being under 400K once the almost-inevitable upward revision happens? The trend is not Obama’s friend here – Tom Blumer has been charting the upward revisions of late, and through the prior 19 weeks, the average upward revision from the initial numbers has been just over 4K. Only 2 of those 19 weeks saw revisions of less than 2K.

The second question is whether it should have been expected. Again, I’ll turn to the seasonally-unadjusted data in the various press releases as that provides the prior year’s data. In 2009, there was a 74K drop in the unadjusted jobless claims between 7/16 and 7/23. In 2010, there was a 91K drop in the unadjusted jobless claims between 7/17 and 7/24. Judging by recent performance, most of this should have been expected.

This post was promoted from GreenRoom to HotAir.com.
To see the comments on the original post, look here.