Keeping in mind the theme of only comparing unrevised figures with unrevised figures, and revised figures with revised figures which I emphasized last week, let's take a look at today's figures:
In the week ending August 18, the advance figure for seasonally adjusted initial claims was 372,000, an increase of 4,000 from the previous week's revised figure of 368,000. The 4-week moving average was 368,000, an increase of 3,750 from the previous week's revised average of 364,250.
No, that's not how you do it! You compare this week's advance number with last week's advance number and you compare last week's revised number with previous week's revised number. It really isn't that hard.
So, the advance number for the week ending August 18 was 372,000. The advance number for the week ending August 11 was 366,000. That means, the advance number went up by 6,000.
Now, the revised numbers run a week behind. Nothing you can do about that. This week's announcement tells us the revised number for the week ending August 11. Last week's press release tells us the revised number for week ending August 4. The difference: 368,000 - 364,000 = 4,000.
I'd say we're headed right back to 390,000–400,000 territory following the weird dip-spike-dip in July, but then I have been known to be wrong.
But, more importantly, I do not understand why none of the qualified data people at the BLS actually say anything about the comparison of advance figures with revised ones.