Despite Weak Jobs Growth, Unemployment Falls Again. Why?

February 10, 2014

The headline job growth announced by the Bureau of Labor Statistics on Friday (link to the report provided above) came in well below analysts expectations at only 113,000 – less than needed to keep pace with growth in the labor force.  Yet, unemployment fell again, from 6.7% in December to 6.6% in January. 

The household survey portion of the report – the portion used to calculate unemployment – was completely out-of-synch with the establishment survey.  That’s not so unusual, but the reason for it was.  Usually it’s because the government relies upon the explanation that more people have mysteriously dropped out of the labor force.  But this time, 523,000 people were actually added back into the labor force, which should have sent unemployment soaring.  Instead, it fell by 0.1% because the “employment level” – the number of people that the household survey says has jobs, inexplicably soared even more by 638,000.  So how do you reconcile those two numbers – a measley 113,000 jobs added by the establishment survey and 638,000 according to the household survey?

Here’s what I believe happened.  At the end of December, unemployment benefits expired for over a million long-term unemployed workers.  When you’re eligible for and collecting unemployment, you must answer “no” on the unemployment form when asked if you’ve found work, in order to remain eligible for your benefits.  If you happen to also be included in the household survey, you’d better answer “no” to that one as well, right?  If you’re collecting unemployment, you’re probably picking up a few bucks here and there babysitting and doing odd jobs, but you’re not going to report that. 

But, once the unemployment benefits are taken off the table, you no longer have that constraint to answering the household survey honestly – a survey which considers you employed even if you earn only one dollar that month as pay for some service performed.  (And it’s a survey for which there was no penalty for dishonestly responding “no.”)  Suddenly, thousands are free to answer honestly that, yes, they made a few bucks babysitting, or helping some contractor friend for a few bucks paid in cash.  Extrapolated to the general population, those now-honest responses add hundreds of thousands to the employment level. 

Regardless of this artificial drop in unemployment, the economy isn’t improving.  Following two quarters of real pre-holiday binge spending-driven growth in the economy, data since the beginning of the year shows an economy that’s faltering once again.  Mortgage applications continue to decline.  Manufacturing is slowing.  New factory orders fell by the most in 33 years in January.  Exports fell as well.  Retail sales growth slowed to zero in the last week of January. 

So January’s employment report is another aberration, on the heels of two other monthly aberrations caused by the government shut-down in November.  Let’s see what happens next month.

January Employment Report: a Home Run or a Foul Ball?

February 3, 2012

The January employment report, released by the Bureau of Labor Statistics this morning, blew past all expectations in every way imaginable.  According to the establishment survey, 243,000 jobs were added in January, far exceeding expectations for 150,000 new jobs.  And, even more impressively, the household survey claimed that the employment level grew by 827,000, and the unemployment rate fell to 8.3%.  It would have fallen even more except for the fact that 508,000 workers re-entered the labor force.  By any measure, this report looks like a real home run for President Obama!

That ball is going … going … going …. Uh,oh.  Did it hit the foul pole?  Was it really a home run or a foul ball?  The report looks pretty rosy until you read the fine print which, apparently, none of the analysts did today.  Near the bottom of the report (link provided above) you’ll find the following note about the establishment survey:

Revisions to Establishment Survey Data

In accordance with annual practice, the establishment survey data released today have been revised to reflect comprehensive counts of payroll jobs, or benchmarks. These counts are derived principally from unemployment insurance tax records for March 2011. In addition, the data were updated to the 2012 North American Industry Classification System (NAICS) from the 2007 NAICS. This update resulted in minor changes to several detailed industries. The benchmark process resulted in revisions to not seasonally adjusted data from April 2010 forward and to seasonally adjusted data from January 2007 forward. Some historical data predating the normal benchmark revision period also were revised due to the implementation of NAICS 2012 and other minor changes related to rounding and the recalculation of aggregate series.

In other words, the baselines for the establishment survey have been changed.  How much?  Who knows?  Nobody.  This was the perfect opportunity to inflate the number of jobs created by monkeying with the baseline. 

And what about the household survey?  What explains the huge leap in the employment level?  Look a little further down on the report and you’ll find the following:

Adjustments to Population Estimates for the Household Survey

Effective with data for January 2012, updated population estimates which reflect the results of Census 2010 have been used in the household survey. Population estimates for the household survey are developed by the U.S. Census Bureau. Each year, the Census Bureau updates the estimates to reflect new information and assumptions about the growth of the population during the decade. The change in population reflected in the new estimates results from the introduction of the Census 2010 count as the new population base, adjustments for net international migration, updated vital statistics and other information, and some methodological changes in the estimation process. The vast majority of the population change, however, is due to the change in base population from Census 2000 to Census 2010.

Once again, with the baselines changed, who knows what the real numbers are?  You could make the numbers say practically anything you want.  If you were Secretary of Labor, appointed by President Obama and, assuming you’d like to hang onto your job for another four years, wouldn’t this be a perfect opportunity to buff up the president’s image with some good news on the jobs front? 

The problem is that this jobs data isn’t corroborated by other economic data.  4th quarter GDP growth was virtually all due to inventory-building.  Retail sales data has been weak in January, following a strong holiday season.  The housing sector is continuing to decline.  The trade deficit is as bad as ever.  The Federal Reserve is so concerned about the weak state of the economy that it recently vowed to keep interest rates at zero for three more years.  Consumer confidence is falling again and the percentage of people saying that jobs are hard to get is rising.  This isn’t a picture of an economy that’s adding 243,000 jobs per month.  Something doesn’t add up.  Looks to me like that ball went foul. 

Nevertheless, here’s my calculation and charts for unemployment:

Unemployment Calculation     Unemployment Chart     Labor Force & Employment Level     Unemployed Americans     Per Capita Employment

* * * * *

If you can believe the data, the 243,000 added jobs break down as follows:

  • Professional & business services:  + 70,000
  • Manufacturing:  + 50,000
  • Leisure & hospitality:  + 44,000
  • Health care:  + 31,000
  • Construction:  + 21,000
  • Wholesale trade:  + 14,000
  • Mining:  + 10,000
  • Retail trade:  + 5,000
  • Government:  no change