Steven HorwitzOne of the distinguishing features of this recession is that the duration of unemployment has been notably longer than in other recessions. See this graph:
This weekend, I've seen a couple of online pieces (second one gated) arguing that one possible explanation for these data is that unemployed folks are less mobile than they used to be thanks to having homes that they cannot sell for a price that comes close to breaking even. [UPDATE: to be clear, I'm not, nor are they, arguing that this is the only explanation for the increased duration of unemployment. Surely benefits extensions matter as well as other factors, but this is one that we perhaps haven't seen before.] Housing policy that subsidized low- or no-equity loans have left a fair number of people in situations where they would like to move to greener pastures in search of employment, but are unable to make the move because they are stuck with houses they can't sell, which in turn prevents them from affording a new home or even a rental in the new location. It's worth noting that the same argument would apply to currently employed people looking for better work elsewhere, opening up their own job for others.
A comparison to Canada is interesting as they did not have the perverse housing policies we had and still have a home ownership rate similar to ours, yet they have not had the fall in housing prices that would limit labor mobility the way that is being hypothesized for the US. The latest Canadian data that I could find has their current unemployment rate at 7.9%, lower than the US, and although their median duration of unemployment in 2009 was 17 weeks, which was slightly higher than the US one was on average, that number was actually just a bit higher than their 2007 and 2008 numbers and the same as 2005. That 17 weeks was also notably lower than their average for the last 15 years. Canadians have always had longer unemployment duration thanks to differing unemployment policies and a weaker economy, but what they have not seen is anything near a comparable jump in duration during this recession. That would appear to be some evidence for the "can't sell the house" explanation for labor immobility and unemployment duration in the US. But I'd love to hear alternative hypotheses.
If this argument is right, it's an interesting unintended consequences story in the way that housing policy interacts with labor market mobility. One can also tell a Lachmann type story here: housing and a job are complementary goods in a household production function. In the face of dynamic change that changes the value of one of the goods, substitution will be called for. Unemployment requires a revaluation of one's human capital. And changes in one's human capital that might make a new job attractive may require rethinking the housing side by moving. But substituting a new job for the old one means that it has to be complementary to the housing piece of the puzzle. In a healthy housing market, swapping out the housing component is not as difficult, which makes it easier for the household to make the necessary shifts in employment/human capital in the face of dynamic change. Where households are stuck with the housing capital, they are also stuck with the range of jobs available in that area. The result is, apparently, longer spells of unemployment and a higher unemployment rate.
I also think this poses some interesting questions for the whole "recalculation" process. Much like the Hoover high-wage policy (though much less dramatic), one of the effects of the housing debacle has been to reduce labor mobility. In the Hoover case, it was a matter of being unaffordable to hire, and in today's case, it's that people can't sell their houses to move to where opportunities might be. In both cases, interventions prevent human capital from finding the places where it is complementary to the new post-crash structure of production. If we want households to recalculate effectively in the face of economic change (e.g., adjust their labor supply when the structure of production changes), we cannot distort the various markets and prices that inform that recalculation process.
If all of this is correct, it would suggest that we're in for a long period of fairly high unemployment until the housing market settles. It may well be the case that some folks who would like to find a (new) job elsewhere will have to wait until they have sufficient equity in their homes to afford to sell. This is yet another reason to get rid of the policies that created this mess so that we don't exacerbate or repeat it in the long run. As for what, if anything, can be done through policy (including, of course, eliminating bad policies) to alleviate this situation in the short run, I have no brilliant ideas. Feel free to suggest some below. The answer, of course, may be "nothing, let the process play out and heal itself."