The relationship between the housing and labor market crises and doubling up: an MSA-level analysis, 2005–2011
A number of studies have examined the recent boom and subsequent bust in the national housing market. Estimates of the housing boom’s start date vary, with suggested possibilities including 1996, 1998, and 2002.7 There is a much narrower band around the date when the national housing bubble burst; the bust occurred in either mid-2006 (estimated using the Case–Shiller Home Price Indices) or in first quarter 2007 (estimated using the House Price Index from the Federal Housing Finance Agency [FHFA]).8 Turning to subnational data, some studies point to considerable dispersion across MSAs in the magnitude of the rise in prices during the boom as well as in the decline of prices during the bust.9 These studies also find a similar set of patterns: (1) MSAs located in the interior of the United States (e.g., Charlotte, Detroit, Cleveland, and Chicago) experienced smaller increases in housing prices during the boom than did MSAs located on the coasts; and (2) the set of MSAs that experienced larger booms also tended to be those which experienced larger busts. Among the interesting exceptions is Las Vegas. In Las Vegas, the rise in nominal housing prices during the boom (150 percent) was not quite as large as that for some other West Coast cities (in both Los Angeles and San Diego housing prices increased more than 200 percent), and the decline in housing prices during the bust was considerably larger than the decline for these counterparts (62 percent in Las Vegas versus around 40 percent in both Los Angeles and San Diego).10
Some studies have also looked at the subnational variation in the timing of housing booms and busts.11 A study by Todd Sinai is the most relevant here, because it looked at both the timing of the most recent housing boom and the timing of the housing bust, and drew comparisons with the housing cycle of the 1980s.12 Interestingly, the study found that the timing in MSAs of the most recent housing bust was more closely concentrated than that of the previous bust, with many peaks around 2007 and 2008 but still a good deal of heterogeneity. Unlike the present article, however, the study did not look at more finely grained (quarterly) data or consider variations in the timing of events in the labor market relative to the housing market by MSA.
Extensive evidence also points to substantial heterogeneity in employment conditions at the subnational level during the latter part of the 2000s. MSAs that were especially hard hit by the labor market crisis, as measured by Bureau of Labor Statistics (BLS) local area unemployment rates, include Detroit, Las Vegas, Los Angeles, and Miami. With the notable exception of Detroit from this set, these MSAs also experienced the housing bubble.13 This is to be expected given the strong relationship between the health of the housing sector and the level of construction employment.14 Although less attention has been paid to regional variations in the timing of employment crises, Howard Wall conducted one such study.15 Wall examined the timing of economic expansions and employment downturns for a small number of cities. He found that these cities experienced these events at around the same time the nation did; however, in line with Sinai’s conclusion regarding variation in the timing of housing crises at the MSA level, Wall still identified quite a bit of dispersion.