Thursday, May 19, 2011

Making claims out of thin air (cont.)

Two weeks ago I argued that the large and unexpected rise in unemployment claims that occurred in April was likely the result of faulty seasonal adjustment factors—which expected a decline in unadjusted claims that failed to materialize—and predicted that claims would retreat in coming weeks. Fortunately, that's exactly what has happened.

The top chart shows the seasonally adjusted series, while the bottom chart shows the unadjusted data. Now that the seasonal dust has largely settled, we see that the supposed rise in claims was a statistical artifact that had nothing to do with what was going on in the economy. Case closed.


Benjamin Cole said...

Good call by Scott Grannis. Kudos.

The bad news is we still have labor participation rates running five percent below pre-recession levels, lots of unused industrial capacity, falling unit labor costs, and now commodities have busted.

The S&P 500 below 1999 levels, and real estate around 2005-6 levels.

I hope we are not sinking into Nippon-itis.

Bill said...


Thoughts on the weak data reports today- philly fed, leading indicators and existing home sales- do you think this suggests a weak 2nd quarter GDP?

Scott Grannis said...

I'm one of those who take the long view of things. I don't believe the economy turns on a dime, and I haven't seen any reason or signs in the data to suggest there has been a fundamental deterioration in the economy. Data can be volatile and misleading. Quarterly GDP growth numbers can also be volatile and misleading, and they can be revised significantly over a period of years, so it's pointless to worry a lot about what second quarter GDP is going to be. I think the economy is still on a moderate growth track (3-4%) and there is a decent chance it could do a bit better than that as the year progresses.