The seasonally finagled headline number for weekly initial unemployment claims was 311,000 last week, soundly trouncing the consensus wild guess of the Wall Street economist crowd, who were looking for 330,000.
Seasonally adjusted (SA) data represents an idealized, fictional, dumbed down, smoothed, abstract impressionistic version of reality that is supposed to more clearly represent the trend. Admittedly, sometimes it comes close, but sometimes it doesn’t.
Rather than using the faked data that the mainstream media loves to idiotically parrot, I like to look at the actual number of claims reported to the Department of Labor (DOL) each week by the 50 state unemployment departments. When that actual data is plotted on a chart with a few filters it’s easy to see the actual trend. It’s easy to see whether there’s any material deviation from the trend in the current and most recent data. It is, in essence, the application of technical analysis to economic data. If you like charts and trendlines, you like this method. If you like abstract impressionist propaganda, stick with the news headlines.
The media ignores it, but the DOL dutifully reports the actual not seasonally adjusted (NSA) data right along with the SA data each week. Here’s what it had to say today about last week:
The advance number of actual initial claims under state programs, unadjusted, totaled 273,411 in the week ending March 22, a decrease of 12,559 from the previous week. There were 315,620 initial claims in the comparable week in 2013.
As a critical observer, I want to know how this compares with the trend, along with how this week’s change compares with the same week last year and with the historical average for the comparable week. It’s like going to the optometrist. Is it better or worse than the trend? Is it better or worse than average? Is it better or worse than this week last year? That information might give us a tip as to whether anything is actually changing, or if job losses are moving along the same curve which they’ve been on all along subject to minor random squiggles week to week and month to month.
