On Structural vs Cyclical Unemployment.

I am extremely reluctant to link to this news story from PBS Newshour.  I do not like to post material for class that I'm not thrilled with since I do not like to write lengthy explanations of the material.  However, I am trying to do two things in this semester.  First, I want to use more video/news in class to illustrate examples.  This was a request made last semester in the course evaluations and I think it makes a certain amount of sense.  Second, I am trying to make more current event material available in general because 1. It justifies my lectures and 2.  I think it's useful in helping you learn the material.  Structural vs cyclical unemployment isn't the kind of story to show up very often in an accessible format, so I am stuck with what I have been given.

Anyway, this video comes by way of Mike Konzcal's excellent blog: Rortybomb:

Watch Can America's Jobless Fill American Jobs? on PBS. See more from PBS NewsHour.

Okay, so, I feel like a major problem with this piece is that it clearly has made up it's mind that the unemployment problem is structural.  I can see why.  The structural story lends itself to TV journalism nicely.  They can go to a couple of locations, interview a couple of managers or owners complaining about the lack of  qualified workers and call it a day.  But it's important to bear in mind that they are presenting mostly anecdotal evidence.  Konzcal, who appears in the story, doesn't get a lot of screen time but he does pose an important counter to the structural story.  As he points out, for the structural story to make sense, you would expect to see unfilled job openings and wages increase.  Neither of which shows up in the aggregate data.

As i said in class I am agnostic on the structural vs cyclical unemployment debate, though I find a lot more to disagree with in the structural position than with arguments for cyclical unemployment.  For instance, the structural story is often told about construction and manufacturing.  However, I feel that if you want to talk about low skilled labor you also have to talk about retail and food service and other sectors that have been on the rise and are also sectors that employ a lot of low skilled workers.  So, lets look at a broader measure of unskilled workers.  Here is a chart that shows the cumulative change from 1980-2012 in employment across four low skill sectors: leisure and hospitality, retail, manufacturing and construction.

I should point out that this is a very rough estimate of the impact of employment changes for "low skill" workers.  Not everyone who works in these industries can be called low skilled.  Anyway, another way to make the point the graph is making is to point out that since 1980 hospitality and retail have added around 11.2  million jobs while manufacturing and construction have lost only 6.4 million jobs.  That total, since its from last month (Jan 2012) is actually kind of distorted since all four sectors have been shedding jobs since the recession.  Before the recession (roughly), from Jan 1980 to Jan 2008 10 million "low skill" jobs had been added.  For the structural story to stick then, you have to to assume we have "reverted to trend" and that around 5 million jobs were not "real" jobs.

Anyway, the point is, in order to tell the story of structural unemployment for low skill workers you have to be able to incorporate the increase in employment in other low skill sectors that are growing.  I have not seen a lot of arguments for structural unemployment that have actually looked at the long run trend in different jobs of different skill level.  So, lets compare job growth in these "low skill" sectors with job growth in "high skill sectors".  Our proxy for high skill will be professional and business services and education and health services.

Here we see much more growth in "high skill" jobs than in "low skill" jobs.  Another way to put this is that  total nonfarm employment has grown by 135% since 1980.  In the high skill industries in the above graph employment grew 85% and in the low skill industries employment only grew by 12% (in 2008 it had grown by 23%).  I want to point out that nothing about that graph suggests that the collapse in jobs created for unskilled workers is structural.  It could very easily be due to the fact that housing is going through a cyclical downturn while the rest of the low skilled service industry is experiencing the same stagnation you see even in the high skilled industries.  This is easily attributable to normal business cycle dynamics.

However, if you wanted to read that graph as a story of structural unemployment then you can tell the story of how the growth of "skilled" employment has slowed down since the salad days of the 1990s but it continues to grow and has recovered more or less to its trend.  Unskilled  labor, however, seems to actually have been on a slight downward trend since its less spectacular 1990s.  Here we can tell a story of the housing boom (and I guess even the tech boom) "artificially" propping up unskilled employment which has now reverted to its pre-1990 level where it "should" be.  Of course, then you have to argue convincingly the definitions of the words in scare quotes.

Anyway, the cyclical vs structural debate tends to run along liberal vs conservative lines.  Conservatives tend to want to frame this as a structural employment problem so they have a reason to throw up their hands and say "nothing can be done" from a public policy perspective.  Roughly speaking, that seems to be the takeaway of the news report above.  However, if the unemployment issue is all about skill levels the ONLY response is a public policy response since the vast majority of the educational system is publicly run.  On top of which, the expansion of education would provide more jobs now for all skill levels.  We would need to hire low skilled workers to build the buildings to house the high skilled workers who should be training workers for the skills they need for a 21st century economy.