That's Odd

If you were looking for signs of stagnation in second quarter growth, you (once again) failed to get it in the latest employment report:

   

   

So where are the skeletons in last month's labor closet?  Most commentators pointed to the failure of construction employment to behave like it should in a good sector in distress.  As Calculated Risk says:

The expected reported job losses in residential construction employment still haven't happened, and any spillover to retail isn't apparent yet. With housing starts off over 30%, it's a puzzle why residential construction employment is only off about 4%.

Macro Man agrees...

While the headline payroll report was as close to consensus as you'll ever see, sure enough there were some revisions that suggested a stronger US labour market that previously thought. Quelle surprise! Perversely, construction payrolls showed another modest 12k gain.

... as does King at SCSU Scholars...

The hard part to figure is how residential construction employment can do so well when all we ever hear is how housing is being hit hard (I've said it myself.)

... who highlights this comment from Real Time Economics (aka The Wall Street Journal econ blog):

The persistent weakness in nonresidential construction payrolls has been every bit as puzzling as the absence of large cuts in the residential sector, though it has not gotten much (any?) fanfare. We continue to believe that the reporting is imperfect and that some housing job losses are showing up in nonresidential while some nonresidential hiring is showing up in residential. –Stephen Stanley, RBS Greenwich Capital

The Journal blog entry also includes a warning about getting too comfortable with recent employment gains:

The behavior of residential construction payrolls remains one of the ongoing mysteries of the monthly employment data, as they are so far out of touch with the behavior of housing construction and home sales that they hardly seem credible… We think an even larger story looming off in the distance will be the annual benchmark revisions to be released in January… It is likely that the annual benchmark revisions will reveal that job growth has been slower than suggested by the monthly reports. –Mission Residential Research

Does such a revision seem plausible?  Well, sure.  The good folks responsible for the ADP National Employment Report claim...

There is a very powerful statistical tendency for estimates of growth of establishment employment, as reported by the BLS after annual benchmarking, to be revised in the direction of estimates previously published in the ADP National Employment Report.

... and year-over-year ADP employment growth has been running below the corresponding BLS payroll rate:

   

   

And if we take Jim Hamilton's advice and average the payroll data with the information gleaned from the BLS' household survey, the picture would look just a little less rosy than it does with the payroll data alone:

   

   

Tom Blumer at BizzyBlog concludes:

June's number was decent.  When combined with pretty significant revisions and no change in the overall unemployment rate, you have the picture of an employment market growing nicely but not spectacularly.

That's a fairly modest assessment, but if you are still a skeptic I guess you are not without ammunition.

Elsewhere: Dean Baker notes that the decent employment growth may not bode well for productivity growth. pgl is unimpressed with reported wage gains.  Max Sawicky wants you to stop looking at the unemployment rate.