For years (indeed, decades) conservatives have claimed that any unemployment that appears to be caused by an insufficiency of demand is really just structural in nature, that the reason that so many people are unemployed isn't because there aren't enough jobs, but because people aren't properly trained for all of the jobs that are available or they simply aren't looking for jobs where they are availabile because they don't live there. Before getting to the fact that such a reality would call for public investment in helping people retrain and relocate as a matter of national interest, which conservatives resolutely refuse, first should come the question of whether such a thing is even possibly true.
Today Paul Krugman, in a pair of blog posts here and here, adeptly addresses the plausibility of such claims from a deeper economic prospective, but there's an even easier way to get at the strict impossibility of conservatives' claims:
That's the number of unemployed people per job opening in the US measured every month. As of the end of last month it appears to be down near 3.5, but that's due in large part to an unexpectedly large number of people ceasing to look for jobs and thus dropping off of the unemployment rolls. Despite the noise in the data, though, it is completely clear from looking at the general trend that the number is well above 2, has been for a long time, and likely will continue to be for years.
The significance of that is the following: If structural unemployment were the entirety of our problem, then there would be at least as many job openings as there are people looking for jobs, which would make that number at most 1. If structural unemployment were even half of our problem, then half of all unemployed people would hypothetically be able to fill available jobs regardless of the means (retraining, moving, etc.), and that number would be at most 2. With that number at almost 4, we aren't even close to a point yet where structural unemployment could have the possibility of being what we should focus on.
Mind the sixty-nining tape worms as we delve into this issue's depths with more graphs...
Now that we've demonstrated the impossibility of the idea that structural unemployment is the source of our problem or should be the focus of our national policy, we can look at the side question of what might individuals do to better their chances in a purely zero-sum sort of way:
That's a graph of unemployed people per job by industrial sector. To clarify things a bit, allow me to split that up:
Here we have Education and Health Services in blue, Finance in red, and Government in green. Note that none of them are near or below 1 (or have been in years), which is where they would need to be in order for people to be able to plausibly switch into the industry because of the promise of a job. Sadly, this is our "good" graph.
Next up is "bad":
That's Retail in blue and Manufacturing in red. There's just no way to construe those as being good enough that we can achieve anything meaningful by ratraining people and moving them around. And that still doesn't address the elephant in the room, our "ugly", Construction:
In brief, yes some industries are doing better than others, but they all are undeniably suffering large unemployment beyond just structural factors, and for most of them a general lack of demand is by far the leading cause of the unemployment they face.