The positive labor market effects of the temporary housing boom 'masked' the negative effect of the decline in manufacturing that otherwise would have been more evident in the mid-2000s.
In Manufacturing Decline, Housing Booms, and Non-Employment (NBER Working Paper No. 18949), co-authors Kerwin Kofi Charles, Erik Hurst, and Matthew Notowidigdo study how two large changes in the national economy during the 2000s affected aggregate non-employment: the continuing decline of the manufacturing sector and the national boom and bust in the housing market. Using detailed data from the Census and the American Community Survey, they estimate that roughly 40 percent of the increase in non-employment over the eleven years between 2000 and 2011 can be attributed to the decline in manufacturing. The decline in employment over that period was largest for men and women without a college degree.
By exploiting variation in housing market dynamics across metropolitan statistical areas (MSAs), Charles and his co-authors find that increases in housing demand sharply lowered non-employment during 2000-7, especially among men and women without a college degree. But the housing market's reversal during the years 2007 to 2011 among cities that had experienced unusually large increases in housing demand during the previous seven years implies that, over the entire 2000-2011 period, local housing booms did not contribute significantly to labor market improvement.
Instead, it appears that the positive labor market effects of the temporary housing boom "masked" the negative effects of the decline in manufacturing that otherwise would have been more evident in the mid-2000s. Thus, the 2007-11 collapse of the housing market not only had an independent adverse effect on labor market outcomes for some sub-groups but also "unmasked" the negative manufacturing effect that would have been apparent earlier. Consistent with this interpretation, the authors use detailed data from the Displaced Worker Survey to show that workers displaced from manufacturing sectors during 2000-7 were much less likely to end up non-employed if they resided in an MSA in which housing demand had increased sharply during the same period.
These results highlight the fact that booms and busts in one sector can have very different aggregate effects on employment dynamics, depending on circumstances in other sectors. In this case, the negative labor market effects of the manufacturing decline are muted during the housing boom and very large during the housing bust.