He writes,

Hurst notes that fewer and fewer Americans with a high school education or less are finding employment in manufacturing. This is a trend that accelerated in the late 1990s. Some of those lost jobs resulted in twentysomethings exiting the labor force. But a great many were absorbed by a thriving construction sector. Between 1998 and 2007, the share of lower-education men employed in manufacturing fell from 15 to 10 percent, virtually a mirror image of construction, where the share increased from 15 to nearly 20 percent.

Read the whole thing. Pointer from Tyler Cowen.

As you know, I am very sympathetic to the view that the period 2000-2010 should be viewed as one of long-term structural decline in the prospects for low-skill males, with the housing boom providing a temporary offset. However, keep in mind two counter-arguments.

1. Overall, the construction sector is very small. I would not want to try to defend the view that the entire boom and bust was in construction.

2. There is still the question of why so much of the lower demand for low-skilled workers translates into reduced employment rather than lower wages. Is it nominal wage rigidity, in which case monetary expansion would provide a cure? Or are there structural factors on the supply side of the labor market that are making dropping out of the labor force more attractive? See the new CBO report on disability. Or, from a different perspective, see Conor Sen on the possibility that leisure has become more highly valued.

In fact, I winced reading this from Fisman:

If Hurst is right, we’re now adjusting to a new normal, one in which there are fewer manufacturing jobs to go around and no housing boom to absorb all the unskilled workers who could have found work in a less globalized and computerized era.

The phrase “fewer manufacturing jobs to go around” is lump-of-work-ish. As an economist, you don’t want to put things that way. Instead, talk about a decline in the relative demand for low-skilled labor, and then talk about how that decline gets divided between lower wages and lower employment.

Note that some people complain that wages of workers at or below the median have not done well. But one can argue that those wages have done too well, and the evidence for that is the huge drop in employment among men without college degrees.

A lot of pundits write as if more aggregate demand would boost both wages and employment among low-skilled workers. However, in standard textbook macro, aggregate demand would increase employment by reducing the real wages of workers.

As you know, I prefer not to think in terms of aggregate demand. Instead, I prefer to think of macro as what Peter Howitt calls a coordination problem, one that I describe as entrepreneurs trying to discover sustainable patterns of sustainable specialization and trade.