Employment and Unemployment

On Friday, April 4 (the first Friday of the month), the March employment and unemployment numbers will be released amid much fanfare.  The February employment change was a negative 63,000, although subject to revision on Friday and again in May.  The January number was a smaller negative, and it too is subject to revision Friday.  The unemployment rate (calculated from a separate household survey) declined in February to 4.8 percent, but the celebration was muted because the decline resulted primarily from a decline in the labor force, presumably related to discouraged workers not job hunting.

I've been honored by an invitation to be a guest host during the second and third hours of CNBC's Squawk Box on April 4, during which time the employment numbers will be released.  I will be on the Squawk Box set if the Dallas rains don't intervene and ground me, as they did last time. 

I'm likely to embarrass myself trying to guess what the employment number will be.  I'll wait until the new claims for state unemployment insurance are released on Thursday, April 3, before formulating my guess, but I imagine it will be a small change one way or another, more likely down than up.

Partly to redeem myself for a bad guesstimate on employment, I hope to have an opportunity to call attention to an aspect of the employment numbers that I wrote about here on July 9, 2007, titled Job Angst: It's the Churn, Stupid.  The point of that posting was to call attention to a little-recognized, but important fact about the job numbers.

My point then, and again now (with more numbers) is that the net changes in monthly employment, even large net changes, are tiny compared to the gross changes in job gains and losses that produce the small net change.  A change of 100,000 net new jobs in March, in either direction, would likely obscure job gains and losses 20 to 30 times larger.  That is an important reason for people's angst about the job situation that seem out of proportion to the size of the reported job changes.  The job churn produces a swift undercurrent even under the best of circumstances.

Let me give you some numbers and calculations derived from a very useful series, but apparently not well known, from the Bureau of Labor Statistics, titled Business Employment Dynamics.  The numbers are for total private jobs only, and thus may not jibe precisely with the numbers you are familiar with.

First, let me make a broad generalization based on those numbers:

In recent years, the quarterly gross job gains and job losses have ranged roughly between 7 and 8 million, or between 2.33 and 2.66 million per month.  Compare that 2.33 to 2.66 million monthly average to the net job gains or losses each month that are usually under 300 thousand.

In 2004, there were 30.9 million private job gains and 28.9 million private job losses, for a net annual increase of 2 million or 171,000 per month.  The gross changes were roughly 14.5 times the net change.

In 2006, a weaker year for net job growth, 30.7 million new jobs were created while 29.1 million jobs were lost, for a net gain of 1.67 million, or 138.6 thousand per month.  That's roughly 16.5 jobs gained and lost for every net job gained.

The even weaker job market in 2007 raises that ratio further.  During the first two quarters of 2007 (the latest data available), 15.2 million jobs were created while 14.5 were lost, for a net gain of almost 700 thousand, or 117 thousand per month.  For each net new job created, over 22 jobs were gained and over 21 were lost.  

No wonder the angst about jobs and the economy seem to bear little relation to the job numbers reported on the first Friday of the month.

It's gross angst.