Unemployment rate goes to 9.0%, only 36K jobs added; Update: 504K disappears from labor force

The jobless rate fell by four-tenths of a point in January to 9.0%, the best downward revision in the post-recession environment, but it didn’t actually mean that the economy added more jobs.  The number of jobs added overall was only 36,000, which indicates that people are still fleeing the job market:

The unemployment rate fell by 0.4 percentage point to 9.0 percent in January, while nonfarm payroll employment changed little (+36,000), the U.S. Bureau of Labor Statistics reported today. Employment rose in manufacturing and in retail trade but was down in construction and in transportation and warehousing. Employment in most other major industries changed little over the month.

The unemployment rate (9.0 percent) declined by 0.4 percentage point for the second month in a row. (See table A-1.) The number of unemployed persons decreased by about 600,000 in January to 13.9 million, while the labor force was unchanged. (Based on data adjusted for updated population controls. See table C.)

Frankly, this doesn’t make a lot of sense. If only 36,000 jobs were added and 600,000 people stopped being unemployed, then the labor force should show a significant contraction. The lower overall rate makes sense if 600,000 people left the workforce, but not if the workforce remained the same. Otherwise, we’d have to conclude that 36,000 jobs represents 0.4% of all employment in the US.

Looking at the A-6 table, which compares numbers January 2010 to January 2011 (not seasonally adjusted), we can see that the unemployment rate for those without disabilities has dropped from 10.4% to 9.7%.  The number of non-disabled adults outside the work force has grown substantially in that period, from 62.8 million to 64.7 million.  That far outstrips population growth and indicates that people are still leaving the work force in large numbers.  In the A-16 table for the same period (not seasonally adjusted), the number of people outside the workforce has grown from 83.9 million to 86.2 million, again showing a large increase.

The topline rate number looks better, but it also looks increasingly irrelevant.  The Department of Labor shows that the average monthly growth of jobs over the last 12 months has been 97,000, not enough to keep up with population growth.  That’s the key measure, and it’s simply not getting any better, nor any more consistent.

Update: Reuters is disappointed … again:

Employment rose far less than expected in January, partly the result of severe snow storms that slammed large parts of the nation, but the unemployment rate fell to its lowest level since April 2009.

Nonfarm payrolls grew just 36,000, the Labor Department said on Friday, far less than the 145,000 increase that economists had expected.

Guess what the Obama administration blames?

The government noted that severe weather could have affected construction payrolls, which dropped 32,000 last month. There were also large declines in the employment of couriers and messengers.

Maybe we need some global warming, eh?

Update II: Zero Hedge notes that the U-6 number went from 16.6% to 17.3%.  He also catches something I missed — the civilian labor force dropped from 153,690,000 in December to 153,186,000 in January after a recent November peak of 153,950,000.  That was a drop of 504,000 in January from December, and 764,000 in two months.  The participation rate dropped from 64.5% in November to 64.3% in December and 64.2% in January.

Update III: Suitably Flip e-mails:

1)  On the 504,000 decline – as you may have seen, that’s due to a Census adjustment:

“The adjustment decreased the estimated size of the civilian noninstitutional population in December by 347,000, the civilian labor force by 504,000, and employment by 472,000; the new population estimates had a negligible impact on unemployment rates and most other percentage estimates.”

The same adjustment decreased payrolls by 472,000 (i.e. the “disappeared” people had an unemployment rate of just 6.3%, so the remaining population should’ve ticked up).  Excluding the Census data, the labor force was unchanged, so I’m still trying to crunch these numbers and get them to agree with the decline in the headline unemployment rate.

The only thing that seems to reconcile the numbers is to (improperly) fail to back out the 472,000, which doesn’t have a “neligible” impact.  It lowers the rate from 9.3% to 9.0%.  With the seasonal and population adjustments, I’m having a hard time making complete sense of it.  If that’s what they did, it was a huge mistake.  Seems quite unlikely, but I don’t see how they get 9.0% otherwise.

And notably, the same estimates that were looking for 148,000 new jobs also expected an increase in the unemployment rate to 9.5%.  That wild disparity (huge underperformance with a significant improvement in the rate) doesn’t make sense, absent a methodology adjustment that’s not accounted for by the Census tweak.

The “adjustment” excuse doesn’t make a lot of sense, either, except that the BLS “adjusted” nearly a half-million people out of the civilian workforce.  Did the US suffer a population decline in 2010?  Flip also decimates the “weather” excuse in this post.