Job Dynamics

Data comes in from an amazing variety of sources telling us just what is up or down in the economy.  If it can be measured, it probably is.  You want jobs information?  There’s information on unemployment, job creation, job destruction, hours worked …  just about whatever you want.  Deciding what is useful is always the hard part – and that is the great skill any of us have to develop in this information age.

A year ago Barataria was focused on Unemployment Initial Claims as a good proxy for total job growth figures that would come out later in the month.  It worked for a while because in 2010 through the summer of 2011 we were at or near the bottom of job creation, picking up steam very slowly.  This figure stopped being a good indicator for reasons that suggest  large employers are still laying off people (those who are eligible for unemployment benefits) while small companies are hiring.

But it would be wrong to drop our study of Initial Claims completely, so let’s take another look.

Below is a chart from our old friends at the St Louis Fed with data on Unemployment Initial Claims since 2000.  A slightly longer view shows how useful this data still is:

Note that in the official “recession” of 2001 more people were laid off than before, hitting 480k per week.  Even in the best of times in the 2000s the level never went below about 300k every week – meaning that in our economy of about 130M jobs about 12% of all jobs are “turned over” every year.  It’s this dynamic nature of employment that makes everything so hard to predict.

But also see how the spike in 2008 hit 640k people losing their job every week – double the level of good times – and has not fallen below 360k despite being on a long slide.  You can look at the trend as good, but it’s hard to say it’s anywhere good enough.

What we do know is that the economy lately has been creating about 200k net jobs per month, on average, since the end of 2011.  That came despite losing about 1,600k jobs, meaning that somewhere every month 1,800k jobs are created – about 1.4% of the labor force are in new jobs in any given month, or about 1 in six workers every year.

That’s a dynamic economy.  We can’t be completely sure that this strong job growth is continuing, but it was the trend for a solid 6 months.

That is what makes the Initial Claims number a bit less useful now.  Every one of those 360k people will tell you that losing a paycheck is a hardship, but overall we have still have some amazing growth.  What keeps that Initial Claims number stubbornly high, though it is falling, is clearly a restructuring as we move ahead to the new economy and leave the old busted one behind.

This is an important point because tackling unemployment is hard enough when 8-10 million people do not have as much work as they want.  We also have what is apparently an ongoing loss due to restructuring that is still shaking out and will continue to do so even as new jobs are being created.

Initial Claims data was useful for predicting jobs with “real time” data collected every week only when we were on the downhill slide from 2007-2010 and into the quiescent period of slow growth in 2010-2011.  It is now a better measure of the old economy fading away underneath us, a process that is not quite over yet.  It may soon get down to the 300k level that approximates stability in the dynamic job market, but for now it is stubbornly higher than it should be at this stage of a “recovery”.

This is simply more evidence that the economic event we are in is not only more severe than anything experienced since WWII, it also has some unusual characteristics.  These characteristics are more typical of what we have seen in past eras that were labeled “Depressions”, marked by very significant restructuring.