More Disappointing News on Pittsburgh’s Labor Market Data

More Disappointing News on Pittsburgh’s Labor Market Data

A month ago, the Department of Labor and Industry released a stunningly negative report showing huge downward adjustments to establishment payroll jobs for 2013—revisions that wiped out all of the significant growth that had been reported each month last year.   In December for example, the so called establishment survey of jobs figure was lowered by 21,000, erasing entirely the originally reported gain.

 

And now, the Department is out with newly revised data obtained through the household survey; the survey that measures the number in the work force, the number of persons employed and the unemployment rate.  Revisions cover all five years from 2009 through 2013 inclusive. Generally speaking the revisions are downward for both employment and labor force and get larger going forward in time from 2009.

 

The revisions are not encouraging.  Remarkably, the downward revision to the number of persons reporting they are working puts the annual average employment for 2013 below the average employment for 2008—instead of the 8,000 gain in employed over the period in the originally reported figures.  Over the last seven months of 2013 the downward adjustments averaged 14,265 per month. Note that for the year as a whole the adjustment was 10,268 lower than the originally reported figures for 2013.

 

Downward revisions were especially pronounced for June and July employment counts with reductions averaging 17,300 per month compared to originally reported figures. New employment estimates for the fourth quarter also showed sizable cuts from original reports that average 13,400 workers for the three months. These reductions put the fourth quarter of 2013 employment nearly 8,000 below the fourth quarter of 2012. Before the revisions, the average for 2013 as a whole stood 8,500 higher than in 2012—after the revisions it was up only 2,300. It is noteworthy that even before the latest round of revisions the monthly gains in late 2013 had already slowed substantially compared to the gains recorded during the midyear months.

 

Fourth Quarter Average Household Employment

Year

Original

Revised

Difference

2008

1,170,103

1,170,103

0

2012

1,175,051

1,169,614

-5,437

2013

1,175,515

1,162,095

-13,420

 

While the number of people working, either self-employed or for another entity, is hugely important as an indicator of economic vitality, it does not by itself however, capture the whole story. Thus, the survey also asks those who report to the interviewer they are not working whether they are actively seeking work. Those who respond “not working but looking” are classified as unemployed.  Added together, the number employed and the number unemployed are called the labor force. Those of working age (over 16) who are not institutionalized and are neither working nor looking for work are termed “not in the labor force”.

 

The unemployment rate, the closely watched monthly statistic that grabs headlines, is simply the ratio of the number unemployed to the number in the labor force usually expressed as a percentage. Interestingly, the number unemployed were revised upward by a monthly average of 527 in 2010 and just under a thousand in 2011. During the first eight months of 2012, the upward revisions in unemployment continued to average around 1,000 but in the latter third of the year, the revisions shifted downward hitting negative 3,000 by year end. This was followed by a 5,890 reduction in January 2013 and further reductions throughout the year. For 2013 as a whole, the revisions lowered the unemployed count by almost 2,000 per month on average.

 

But as it turns out, the unemployment rate itself was barely changed by revisions for most months over the five year period except in the few months with exceptionally large revisions in the number unemployed.  Note that in both the original reports and the revised data, the number unemployed began to fall rather sharply in the second half of 2013.  At the same time, employment began to fall below the twelve month earlier levels reflecting a distinct weakening in the labor market.  So, of necessity, with both unemployment and employment falling the labor force also fell and by more than employment to reflect the decline in both.

 

What does this tell us about the recent dramatic declines in the unemployment rate?  A drop in the number unemployed accompanied by rising employment would be unambiguously positive in that it would point to some unemployed persons finding work. However, with employment falling or holding steady when unemployment falls, the implication is that out of work persons are not moving into jobs but rather out of the labor force altogether, that is to say they are no longer looking for work.  Some argue that the falling labor force could reflect retirements.  That might be possible but it would suggest that as people with jobs are retiring some of the unemployed are filling the positions opened up by the retirements including those created by promotions into the vacancies resulting from the retirements. It could happen, but the sudden rather rapid decline in unemployment in the second half of 2013 implies something else is at work, either in measurement issues or in actual changes in labor market dynamics.