Good labor stats, seasonally adjusted for your comfort
The BLS reports today that the job picture in the US is looking pretty good. After a very disappointing May and June, July turns in a better-than-expected number.
Nonfarm employment grew by 207,000 in July, and the unemployment rate was unchanged at 5.0 percent, the Bureau of Labor Statistics of the U.S. Department of Labor reported today. Over the month, payroll employment rose in many service-providing industries.Note that the Bloomberg consensus forecast was just 175,000 jobs. With July plus a 20,000 job upwards revision to the weak June numbers, The Capital Spectator says we're "cookin' with gas" now.
I can't help but think that Barry Ritholtz is basically right, however, when I look at the seasonally unadjusted numbers. Barry is always going on about the funny number crunching done by the BLS among others, and the July figures seem out of sorts to me.
Case in point: the SA jobs number for July was +207,000 while the NSA number was -1,201,000. That in and of itself is not what caused my eyebrows to rise. What did is the fact that the NSA figure for July 2005 (-1,201,000) showed greater job loss than in July 2004 (-1,143,000), and yet the SA figure for July 2005 (+207,000) is two-and-a-half times larger than the SA number for July 2004 (+83,000). This is quite unlike January and April of both years, in which the SA and NSA figures are very similar (SA: 1/2004, +117; 1/2005, +124. NSA: 1/2004, -2661; 1/2005, -2692 . . . SA: 4/2004, +337; 4/2005, +292. NSA 4/2004, +1131; 4/2005, +1181). So why are the SA July figures so dramatically different from 2004 to 2005 when the NSA figures are pretty similar -- and especially since the NSA figures for 2005 are worse than 2004?
My second concern: if we're now "cookin' with gas", why is the NSA job total for January through July 2005 (+344,000) actually lower than in 2004 (+358,000)?