Does more homeownership lead to higher unemployment?

REUTERS/Rick Wilking
Owning a home has over the years become almost synonymous with achieving the American Dream, and we have embraced the idea with near-religious fervor.

Here's a new study that your real estate agent would probably swallow or burn rather than show you.

Two economists argue that high levels of homeownership, to quote one of them, Andrew Oswald of Warwick University in England, "mess up the labor market." (Actually, it was "labour," but let's not hold that against him.) Using all kinds of data on 2 million randomly selected Americans and a number of equations I can't begin to understand, he and his confederate, David Blanchflower of Dartmouth, document "a strong statistical link between high levels of homeownership in a geographical area and later high levels of joblessness in that area." The word "later" refers to the authors' contention that the ill effects can take up to five years to show themselves.

Zoom through all the analysis, and you'll find that the authors of the study, which was released by the Peterson Institute of International Economics, a nonprofit, nonpartisan research group (best known for its deficit hawkery), are homing in on the fact that the five states with the largest increase in homeownership from 1950 to 2010 — Alabama, Georgia, Mississippi, South Carolina and West Virginia — had a 2010 unemployment rate that was 6.3 percentage points higher than in 1950. The unemployment rates in the five states where homeownership rose the least — California, North Dakota, Oregon, Washington and Wisconsin — rose 3.5 percentage points during the same period.

What about Minnesota?

I asked Blanchflower to explain why Minnesota, with its long-standing high rate of homeownership (now 72.6 percent) and relatively low unemployment rate (5.3 percent), didn't conform to the theory; his reply was an email snort to the effect that I didn't get econometrics. (No kidding.) Alabama's rise in homeownership over the past 60 years was 20.3 percent while ours was 6.6 percent, and Alabama's unemployment rate is a couple points higher than ours.

Homeownership is pretty high in the United States. I wanted to see whether the authors' thesis held up in European countries where renting is more common. So I devised my own simple-minded non-econometric table.  

CountryHomeownership rateUnemployment rate 2012
Slovakia97.014.0
Romania96.07.5
Estonia96.010.2
Bulgaria95.612.3
Malta95.06.4
Hungary92.010.9
Lithuania91.013.3
Croatia90.115.9
Spain85.025.0
Latvia84.914.9
Slovenia83.08.9
Ireland79.014.7
Norway77.03.2
Portugal75.015.9
Greece74.024.3
Luxembourg70.05.1
Italy68.510.7
Belgium68.07.6
Cyprus68.011.9
United Kingdom66.47.9
Finland66.07.7
United States65.48.1
Poland62.410.1
Japan60.04.3
Turkey60.08.1
Netherlands58.05.3
France57.010.2
Austria56.04.3
Czech Republic47.07.0
Denmark46.07.5
Germany42.05.5
Sweden41.08.0
Sources: Eurostat; 2012 Housing Europe Review

The theory seemed to hold up, although there were anomalies. Norway, for example, has an enviably low unemployment rate and high homeownership, Blanchflower explained, because "it's oil rich." Bulgaria may have an artificially high rate of homeownership because governments of former Soviet states and satellites practically gave their public housing away after the fall of the U.S.S.R. (And I can't quite bring myself to believe that Bulgaria's real unemployment rate isn't way north of what the government is reporting.)

Similarly, North Dakota employment is booming right now, not because few people own houses, but because there's been a discovery of a new natural resource.

Nevertheless, Oswald and Blanchflower say that the pattern holds even when they control for other elements — and, most important, when they filter out the last five years of housing market collapse.

Oswald has been drumming on this theory since the mid-1990s, when he compared (in an econometric way, of course) homeownership and unemployment in European countries. But back in 2002, the National Bureau of Economic Research contended in its own paper that homeowners were not more likely to be unemployed than renters. In fact, the opposite is true. Terrified by the prospect of foreclosure, homeowners historically have been more likely to take any job than hang around the house watching daytime TV. Anyway, as a result of the NBER's findings, academics pretty much  shelved the Oswald thesis.

'Negative externalities' 

In the new paper, Oswald and Blanchflower concede that homeowners are not prominent among the ranks of the unemployed. They insist instead that that high rates of homeownership create negative externalities that dampen the labor market "like a treacle blanket thrown over the surface of the country and the economy," says Oswald.

Negative externalities are collateral damage that affects a third party who gets stuck with the costs. Example: Let's say that I have a factory that produces shoes with 7-inch stiletto heels, and every woman in town buys them. Those spiky shoes, as it turns out, dent every wood floor their owners walk on. Neither my business nor consumers have paid (or probably will pay) for the repairs. Unless somebody sues me, the floor-owners have to eat them. More common externalities of economic activity are pollution and congestion.

So what are the negative externalities of high rates of homeownership? For one, it may induce immobility. If you own a house, you may not be unemployed (or stay that way for long). But you could be less inclined to move to take a new job, even a better one. Houses are not liquid assets; selling one entails a ton of work and a raft of transaction costs. (And, of course, people often love their houses, despite their leaking roofs and endless maintenance requirements.) That unwillingness or inability to budge vitiates economic growth.

Other arguments? Places with high rates of homeownership also have longer commute times. To be able to buy a house, many people have to live further from their jobs in town centers. And, if they change jobs, from, say, one in Robbinsdale to one in Bloomington, they can't exactly up and sell their home in Robbinsdale. Instead, they wind up on the road. The longer commute times produce more congestion, which ramps up costs for employees and employers who then are less likely to add more workers.

Finally, the authors show an association of high homeownership rates and lower rates of business formation. "This might be the result of zoning restrictions and NIMBY effects of a kind that...are rational for homeowners," the authors conjecture. Not-In-My-Back-Yardism also discourages construction of rental housing or any kind of structure that doesn't conform to what's there already.

My econometric training (none) is insufficient to tell you whether Oswald's and Blanchflower's association between the labor market and homeownership is strong or merely incidental. But they are probably correct in asserting, as Oswald has,that "modern generations have been brainwashed to believe there is something wrong with them if they rent."

Owning a home has over the years become almost synonymous with achieving the American Dream, and we have embraced the idea with near-religious fervor. And although in the aftermath of the housing collapse there's been an uptick in people expressing a preference for renting, it's almost unnoticeable in the avalanche of still strong sentiment for homeownership. In a 2012 survey conducted by Fannie Mae, for example, 70 percent of renters said that homeownership is a better alternative than renting; and 63 percent  of them have aspirations to be a homeowner someday.

To hear the National Association of Realtors tell it, homeownership has only positive externalities. A study the group distributed claimed that homeowners' children did better in school and were less likely to drop out, that they were better behaved and that teen pregnancy rates were lower than those of renters' children. Homeownership (or housing stability) lowers crime, creates health benefits, increases participation in community life and leads to better upkeep of property. The only claim they didn't make was that homeownership  lowers your BMI.

If it's true that high rates of homeownership stifle the labor market, however, maybe in the interest of creating more jobs, we should curb some of the tax benefits of homeownership, like the mortgage interest deduction, and make renting a little less unattractive.

But, with all the sentiment (or sentimentality) about homeownership surging through our veins and the hard commercial interests of the real estate industry, we'll probably remain sold on the idea that homeownership is only one step away from godliness.

Comments (5)

  1. Submitted by Gerald Abrahamson on 05/23/2013 - 09:34 am.

    You missed the point of the article.

    The authors were talking about *significant increases* in home ownership (as a historic ratio).

    Economies develop and grow based on the current economic situation–which reflects home ownership. Thus, the businesses that get created are based on current local (non-)home ownership. Thus, there would be fewer (smaller) home ownership oriented businesses (per capita) in an area with low home ownership–and more (or larger) home ownership oriented businesses (per capita) in an area with high home ownership.

    The point made in the cited article is a CHANGE from being a (relatively significantly) low home ownership area to being a historically (relatively significantly) high home ownership area. Going from LOW to HIGH home ownership ratios implies a higher unemployment rate.

  2. Submitted by jody rooney on 05/23/2013 - 09:48 am.

    A simple regression on the numbers

    that you posted for Europe gives an R squared of about .17 unadjusted. While an R squared of .50 is pretty good for cross sectional data .17 seven is not all that great, meaning as a single variable ownership does not explain a lot of the variance in the unemployment numbers.

    So at least we can say home ownership does not correlate with very well with unemployment.

    But these folks are talking about the change in ownership not absolute ownership. I think one might also want to through in the change in the housing stock as a variable. If housing construction is cyclical based on the demand for ownership this might be a self fulfilling prophecy. That is: As demand for ownership rises more houses are built employing more construction workers, as the supply catches up with demand then there is an oversupply of houses resulting in unemployment for construction workers and their suppliers. Because construction may be a large part of the economy then you have an increase in unemployment.

    There are significant differences in the economies of the ten states identified in the article which could have a bigger correlation with unemployment than changes in home ownership.

    Econometrics is basically a sophisticated playing with the numbers. So you could square the ownership value and add it to the equation, or use logs but you can over analyze the data beyond what the data actually says. Something that came up during my orals because I minored in statistics. I would have to look at their work but I think these folks are reaching.

    The biggest concern is that correlation is not causation and sometimes you need to dig a bit deeper for the answer.

  3. Submitted by bruce fisher on 05/23/2013 - 11:43 am.

    Homes/jobs

    Home ownership could still work and not suppress jobs, if a homeowner taking a new job in a different town would rent their house to some home owner that had secured a new job in that town, and the rent a house in new town they secured a job.

  4. Submitted by Connie Sullivan on 05/23/2013 - 04:58 pm.

    Aside from the correlation/causation problem here, there is a real lack of context of people’s lives in this speculative, numbers-and-formulae-based exercize. Dancing in the air, pure game.

    And, what is the point? That people not buy homes, fearing they’ll make the local economy worse at least in terms of unemployment? That people simply relax and not feel ashamed [?] about the fact that they rent their dwelling? That people should not buy houses so they can move to XXX to capture a job (we can all go to Bangladesh, for example, or at least to Guatemala, where there are lots of ill-paying jobs to be had).

    What’s wrong with this elaborate and irrelevant econometrics study is that it’s almost impossible to get from it either to a social reality or any policy direction.

Leave a Reply