The usually-reliable David Leonhardt gets it wrong, I think, when he writes:
Arguing Over the Jobs Slump: I agree that structural unemployment is a major problem. You can see it in the fact that the unemployment rate for less educated workers has risen much more than for more educated workers...
But unemployment always rises more for less educated workers in recessions and falls by more in booms:
More educated workers have bigger and better job-search networks, and the same things that made them more educated also make them make better use of their networks. When the labor market goes south, the consequences are much worse for those for whom the system does not work terribly well in normal times. But that doesn't mean that any significant component of unemployment is "structural."
Unemployment is "structural" when attempts to boost spending boost not employment but rather prices because the products where demand increases are not products that can be made by hiring the currently-unemployed. You see structural unemployment when there are significant groups of businesses and industries that are frantically raising wages in an attempt to attract more qualified workers while wages in the economy as a whole are stagnant.
We don't see any of that.
There is no date currently showing that structural unemployment unamenable to cure by spending stimulus is any significant part of our current problem. (Of course, we will if unemployment stays near 10% for very much longer: cyclical unemployment turns into structural unemployment.)
As Scott Sumner writes:
January 2006 — housing starts = 2.303 million, unemployment = 4.7%
April 2008 — housing starts = 1.008 million, unemployment = 4.9%
October 2009 — housing starts = 527,000, unemployment = 10.1%
[H]ousing starts fall by 1.3 million over 27 months [from 1/2006 to 4/2008], and unemployment hardly changes. Looks like those construction workers found other jobs, which is what is supposed to happen if the Fed keeps NGDP growing at a slow but steady rate. Then NGDP plummeted [from 4/2008 to 10/2009], and housing fell another 480,000.... [T]he huge run-up in unemployment did not occur when the big fall in housing construction occurred, but much later, when output in manufacturing and services also plummeted.