A number of sources have cautioned that the seasonal adjustment was behind last weeks goof New Claims report and that we should be cautious.

We should always be cautious yes, and the moving averages do a lot better than the week to week and even then trend spotting takes patients and humility.

However, I don’t think we should get too worked up over it. Check out the graph of seasonally adjusted versus not adjusted numbers.


There are a couple of things to look at. One, if you look at the green arrow you can see that in the late summer of last year unadjusted claims were falling but the adjustment said things were getting worse. Just like now its because unadjusted claims were moving in the expected direction fast enough.

The story told by the adjustment turned out to be correct.

Two, the pair of black arrows show that unadjusted claims deviated from adjusted claims in opposite directions in May and June.  In May Chrysler and GM had early shutdowns and so New Claims didn’t fall as much as the seasonal adjustment expected. The seasonally adjusted trend flattened out and caused a lot of heartache over “withering shoots”

However, that also meant less need for standard retooling shutdowns in June and so that layoffs are now rising less than expected. The net effect has been to send the seasonally adjusted numbers down to a level more consistent the pre-stages of recovery.

We can’t know for sure, but my sense is that taken together these two moves about cancel each other out and the seasonally adjusted number is probably on track.