Known for their seminal work in the economics of happiness, David Blanchflower of Dartmouth College and Andrew Oswald of the University of Warwick have been the center of a big media splash over the past week. Their most recent paper, entitled "Does High Home-Ownership Impair the Labor Market," has legions of anonymous internet commenters and Op/Ed articles alike chastising them for their hasty conclusions. But perhaps they're not incorrect, only misunderstood.
It seems irreverent to suggest that homeownership actually increases unemployment. After all, the New Deal-era Federal Housing Adminstration (FHA) extended the mortgage loan period and introduced amortization to drive economic growth and consequently, employment. It's one of the few politically non-partisan issues today: Homeownership is an intrinsic part of the American Dream.
However, Blanchflower and Oswald do not contend that homeowners tend to be unemployed themselves; instead, they suggest that homeownership has spillover costs that extend into the labor market. Using data collected from 1986-2011, one of their arguments centers on how residential sprawl tends to stifle the growth of new business and job creation. Apart from the obvious zoning issues, there will always be suburban "nimbies" (not in my backyard) protesting the new McDonald's opening near the local dog park. Others have noted that such results may reduce the stigma of removing mortgage tax subsidies; indeed, the benefits would proliferate in a multi-tiered fashion.
The paper further forwards the proposal that commute times suffer from homeownership. The casual reader may cast a skeptical eye over the implicit association here -- that the relatively minute time and congestion costs borne by commuters are somehow inimical to employment prospects. However, recall that the homeowners themselves are not the ones predisposed to suffer from unemployment. Instead, a new class of "backyard employees" face commutes that raise area costs for both executives and workers, dampening the region's economic prospects as a whole.
To tie this all in, they note that there is a strong time-lag element in the correlation between homeownership and unemployment rates. In other words, unemployment rates only seem to be associated with an increase in home-buying three to five years prior. It could be that this delayed manifestation is due to the spillover nature of the costs of homeownership.
This is the argument most prominently supported by the data. The relationship is strongly significant, and the magnitude of the relation increases as lag increases. Perhaps their model can explain the association between recent state-wise spikes in unemployment and the residual effects of the 2006 housing bubble.
The two researchers take plenty of precautionary measures to ensure that their results are not "accidentally" true because of differences between states (their basic unit of analysis). For example, job creation in a sparsely populated and resource-rich state like Maine may remain strong regardless of whether tenants own or lease the land. They employ special kinds of indicators called state-dummy variables to capture and account for these effects. Many of their results remain significant even after controlling for these state-specific institutional differences.
Blanchflower and Oswald ultimately conclude that the relationship between homeownership and unemployment has "elasticity" greater than 1. This is economic jargon which indicates that a rise in homeownership causes a more than proportional rise in unemployment -- a truly startling finding.
Such research yields new and surprising insights into relationships that were heretofore unknown, but it should not be taken as a new battle cry for policymakers. Even at the macroeconomic level, unemployment targeting is not the solitary goal of monetary policy.
"For a family, a home is generally its most significant asset and serves as its primary wealth-building vehicle," said former Federal Reserve economist Edward Gramlich in a 2001 address. "Survey results consistently indicate that buying a home ranks among the top motivations for saving."
Policy design may be the public face of economics, but it is one of the most tenuous aspects of the field. It requires garish speculation, with politics and emotions filling in the quantitative gaps. No one will revive the flailing labor market by overhauling the keystone of American culture that is homeownership, but decision-makers must internalize the conclusions of this paper for what they're worth.
The authors concede that they were not able to control for changes in the basic framework of the housing market, as this would have required more sophisticated techniques than just dummy variables. University of Tampere economist Jani-Petri Laamanen did take this into account in a parallel study on Finland, though. He obtained results that corroborated the outcomes obtained in this paper. Still, the question has much research potential, since the Finnish housing market is undoubtedly influenced by different factors than the U.S. market. The work has only just begun.