Thursday, May 5, 2011
The news that first time claims for unemployment surged by almost 90K during the month of April has rattled investors' confidence. But a look at the underlying data suggests that the source of the unexpected rise was most likely a faulty seasonal adjustment factor.
The top chart shows the level of non-seasonally adjusted (i.e., actual) claims. Here we see only a very modest uptick, nothing unusual. If that were all you knew, you would say that absolutely nothing happened in the labor markets in April. The second chart shows the level of seasonally adjusted claims, with a huge spike. Yikes! But the huge spike didn't occur in reality, it was entirely the result of applying seasonal adjustment factors to the raw data. Sometimes the seasonal factors can fail to anticipate the reality (i.e., the adjustment factor expected claims to fall meaningfully in April, but they didn't), and this is probably one of those times. If so, then we should see claims fall back down in the weeks to come, and I expect that will happen.
Update: Here's a chart of non-seasonally data showing how many people currently are receiving unemployment compensation of some sort. Once again, we see that there has been no sudden deterioration in the real world. All measures of those receiving insurance have been declining of late.
Posted by Scott Grannis at 9:10 AM