Today, seasonally adjusted claims for unemployment fell "unexpectedly," and the market is responding positively, just as I have been predicting for weeks. But in fact, all that has happened is that not as many people are getting laid off in the auto industry in July as the seasonal factors expected. That's because they were laid off last April instead.
The first chart above shows the seasonally adjusted level of claims, which have declined in July. The second chart shows the actual level of claims, which for the first time this year has risen meaningfully. Seasonal factors expected claims to rise even more, much as they did at this time last year, so they subtract from the actual number.
Whatever you might think is actually happening to the real economy, what this shows is that seasonal adjustment factors can actually be wrong, in the sense that the economy doesn't always behave as expected. Sometimes you have to be skeptical of jobs numbers (in particular the June employment numbers released last week). The market and the headlines are somewhat excited about today's claims number, but I'm not. I think it's just payback for the "unexpected" increase in claims last April, when auto layoffs happened earlier than expected. The reality is that the economy remains in a slow-growth recovery, and any pickup that is suggested by the claims numbers is illusory.