Greg Mankiw looks at unemployment and sees something scary. I agree:
He observes :
“This recession looks very different, and much more troubling, than those in the recent past.”
What is different? Lots of things certainly, but look at employment in the construction industry:
From the mid-1990s through 2006, construction employment accelerated and grew much more quickly than the previous trend. You can see previous recessions in the construction employment data–it takes anywhere from 2-3 years for construction employment to bottom out and start recovering. That has yet to happen in this recession and it’s been about four years of declining employment in the construction sector.
Wonder why? Look at this graph of housing starts–housing under construction. I’ve narrowed it to starts of single units–houses rather than apartment buildings:
From the mid-1990s through 20006, there was an unprecedented (at least in the modern data) expansion of home construction. (I explore the role of public policy in this expansion here  and here .)
Is this a cause or an effect of the troubles in the labor market? I’d argue it’s a cause. The Wall Street troubles are the effects of this collapse though Wall Street helped cause the problem as well.
You can see two of the last three downturns in these data (though not the mild recession of 2001.) In each case, housing starts rebounded in less than three years. The current recession has seen four years of decline in housing starts of new single-unit homes. Why? Two reasons. One is that it’s a particularly bad recession. But the second reason is that policy-makers have worked tirelessly to prop up housing prices. They did that for many reasons, some attractive, some not so attractive. But one of the effects has been to keep housing prices artificially high, keep the stock of vacant houses artificially high and discourage new housing construction.
As I have written before (echoing Arnold Kling on recalculation), if you are a construction worker who builds residential homes, it’s not clear whether you should wait and hope the housing sector bounces back (and you will be able to use any specialized skills you’ve accumulated) or switch sectors in hopes of finding a job that will probably pay less than construction.
Lessons? Maybe Ed Leamer was on to something important  and Keynesian aggregate demand stimulus is not going to help unemployed construction workers until the housing market is allowed to return to normal pricing and quantity responses. And I would mention once again that 25% of the job losses since the onset of the downturn have been in construction.