Last week we declared watching the initial weekly unemployment claims numbers was just nuts. This week we've upped it to super nuts. Why?
Because the numbers are then revised after the press release and there is so much noise in these statistics one cannot jump for joy (or cry in your beer depending) on these reports weekly press statements.
Take this week for example.
In the week ending Sept. 5, the advance figure for seasonally adjusted initial claims was 550,000, a decrease of 26,000 from the previous week's revised figure of 576,000. The 4-week moving average was 570,000, a decrease of 2,750 from the previous week's revised average of 572,750.
The advance seasonally adjusted insured unemployment rate was 4.6 percent for the week ending Aug. 29, a decrease of 0.1 percentage point from the prior week's unrevised rate of 4.7 percent.
The advance number for seasonally adjusted insured unemployment during the week ending Aug. 29 was 6,088,000, a decrease of 159,000 from the preceding week's revised level of 6,247,000. The 4-week moving average was 6,182,500, a decrease of 37,750 from the preceding week's revised average of 6,220,250.
The fiscal year-to-date average for seasonally adjusted insured unemployment for all programs is 5.617 million.
Sounds good right? Dramatic decrease? Well, last week's report makes the decline 20k, not 26k initial claims. Then because the moving average was revised, they report a 4 week moving average decrease. But last week's moving average was 571,250, so they revise the moving average upward and wala, it looks more like a decrease.
This is why one must take a lot of data points on unemployment statistics and not go nuts over one week's report. The absolute numbers on initial weekly unemployment claims is still in deep recession turf. That's the bottom line here.
Or borrowing Economist Brad Delong's graph, we should reiterate his conclusion on the overall absolute levels, sigh.
Src: Brad Delong, click on graph to enlarge