Back on July 24th, I had said:
In the week ending July 14, the advance figure for seasonally adjusted initial claims was 386,000, an increase of 34,000 from the previous week's revised figure of 352,000.
… I think the odds are better than 50-50 (based on nothing other than gut feeling), that the report for the week ending July 21 will show an upward revision for the July 14 figure, meaning the actual increase in initial unemployment claims for 7/15–7/21 will be higher than the 34,000.
In the week ending July 21, the advance figure for seasonally adjusted initial claims was 353,000, a decrease of 35,000 from the previous week's revised figure of 388,000.
Yey! I was right. The real increase for the week ending July 14th was 36,000, not 34,000.
Now, let's look at the August 2 press release:
In the week ending July 28th the advance figure for seasonally adjusted initial claims was 365,000, an increase of 8,000 from the previous week's revised figure of 357,000.
The revised figure for the week ending July 21 was revised upward to 357 thousand, from 353 thousand. That means, the decrease reported, 35,000 was actually overstating the decrease in the initial claims, and therefore, the decrease should have been 31 thousand.
Of course, the revision also made this weeks increase look smaller. Had the previous week's figure not been revised, initial unemployment claims this week would have gone up by 12,000.
But, my problem is not with revisions (which are a necessary evil presumably because of delays and errors), and seasonal adjustments, without which we could not make week-to-week comparisons (seasonal adjustment is always hard, but I have no reason to think there is anything fishy going on).
Nah, the real problem is how these figures are reported. So-called journalists seem to pick and choose whatever number is less damaging to President Obama. Is the number less than previous week's revised number? Make sure everyone hears about it. Is the number higher than previous week's but less than some panel's expectations? Well, then, make sure that's in the headline or first paragraph.
But, the real problem is, one can get whiplash trying to follow this week by week stuff.
And, comparing with the same period last year is always a good way to gauge if more and more people are filing initial unemployment claims.
The data are available on the Labor Department's web site.
Here are the not-seasonally adjusted (but, presumably revised) numbers for the end of July:
07/31/2010 402,140 07/30/2011 341,103 07/28/2012 310,492 (from press release)
Does this mean things are booming?
According to the Bureau of Labor Statistics, 143.2 million people were employed in June 2012 which is about 3 million higher than June 2011's figure of 140.1
Over the same period, the civilian, non-institutional population of the U.S. went from 239.5 million (total about 309 million) to 243.2 million (who knows what the total is).
That is, the proportion of those employed out of the civilian, non-institutional population went from 58.5% to 58.9% between June 2011 to June 2012.
Here's what's been going on with that: