The Sky Is Falling. And Other Jobs Report Observations.

We’re on a bit of a role here talking about big-picture economic stuff, and today’s going to be yet another.  You might have noticed that Friday the government issued the “Jobs Report” – you know, that report that talking heads and pundits get breathless over every month.  This past month there was building anticipation for a report indicating growth in the range of 260,000 net jobs. Instead, the report indicated 142,000 added jobs.  Holy cow! The sky fell!

Maybe.  But why this intense focus on month-over-month data?  Is it really telling us all that much?  As a directional indicator, maybe it is.  But I’d posit that a longer view is probably far, far more useful.  Not only are there cut-off issues and seasonal issues and an imponderable number of other variables, but the actual data collection process is a bit zaney.

What is the Jobs Report?

The “Jobs Report,” as it is commonly known, is officially the Employment Situation Summary released by the Bureau of Labor and Statistics.  The report is principally derived from two things: the household survey and the establishment survey.  Both are statistical based collections from which the BLS estimates overall numbers.  For instance, the household survey consists of an actual survey of 60,000 households.  The BLS outsources this job the the U.S. Census Bureau.  This survey is both a telephone survey and a good old, door knocking, “Hi I’m from the government can we chat?” survey.   On the other front, the establishment survey is a sample of approximately 145,000 businesses.

Based on this two pronged approach, the BLS then publishes their report on the first Friday of every month.

 

Revisions (Revisions, and More Revisions)

Would it surprise you to know that the Jobs Report, the one we breathlessly wait for, is not the final report?  There are actually 2 official revisions after the the release.  What’s that, you say? They probably aren’t a big deal?

Behold: Here is a chart reflecting the report revisions between the initial release and the third (final) release going back to 2002.  And yes, you are reading that correctly: a revision of 250,000 jobs.  Assuming a 42,321 average annual salary, that annualizes to a $13.3 billion oops.

job_report_revision_thousands

But maybe that’s not fair.  What is those months has big absolute numbers and the revision was actually just a small percentage of the whole revision.  THEN we’d be justified in hitting refresh every few minutes on those release days.  But alas…

job_report_revisions_percent

Take Aways

So basically no matter how you look at this…. the revisions are an important part of the report.   In fact, when you drill down on the reporting procedure it turns out that only about 80% of the data is actually collected by the time the initial report is issues.  What are we getting at?  Well, clearly the employment situation is an important metric and gives a directional indicator of how the economy is trending, the idea of getting caught up in this monthly release is a bit contrived.  As with a lot of things, let’s expand our horizon a bit and look at longer term metrics before yelling that the sky is falling.

 

P.S. – If you want an even fancier visual check out this from Bloomberg.


Sources

BLS Current Employment Statistics Report (CES)

More details on where the CES Data comes from

National Average Wage Index from SSA

Most Recent Report Release

Some great FAQ’s from BLS on The Employment Situation

The source data for our charts.  Charts are home brewed.  We pulled ours from the the CES Databases

 

 

 

X