How bad is the employment picture, really?

Could it be worse than I thought? We quite likely have not hit bottom in the labor market yet, and the percent loss in nonfarm payroll employment since the beginning of the current recession is already worse than all of the previous seven recessions save the 1981–82 contraction:

6a00d8341c834f53ef011278d658e128a4-400wi

I thought that was not so great, until I took a look at Spencer England’s chart posted by Barry Ritholtz at the Big Picture. (The graph had shown up earlier at Angry Bear and was noted in turn by William Polley.)

Like my graph above, Spencer England’s provides a cross-recession look at employment losses, but based on data from a survey of households (as opposed to payroll data collected from business establishments). Here’s my version of that chart:

6a00d8341c834f53ef011278d6591028a4-400wi

From that look, the labor market in this recession is off-the-charts bad.

There are several reasons the payroll and household employment statistics might differ, and I was puzzling over them when Menbere Shiferaw, one of the many ace analysts here at the Atlanta Fed, came to my assistance. Deep in the details of the latest Bureau of Labor Statistics employment report is this:

“Effective with data for January 2009, updated population estimates have been used in the household survey… 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 primarily from adjustments for net international migration, updated vital statistics information, and some methodological changes in the estimation process…

“Data users are cautioned that these annual population adjustments affect the comparability of household data series over time. Estimates of large levels such as total labor force and employment are impacted most.”

So it may not be such a good idea to use the household employment data to benchmark the job picture with past recessions. And I think I’ll stick with the payroll series for historical reference. Nonetheless, I think we can readily agree that both series are giving us a similar message about the labor market in the here and now, and that message isn’t a good one.


Originally published at Atlanta Fed’s Macroblog and reproduced here with the author’s permission.

One Response to "How bad is the employment picture, really?"

  1. Guest   February 17, 2009 at 2:39 am

    Sir:Could it be the household survey includes all of the folks on severance as well as consultants and part-timers who aren’t included in the labor stats?