Back in the 1990s, British economist Andrew Oswald first showed that higher levels of homeownership were correlated with higher levels of unemployment across European countries and within the United States. Other possible variables — such as unionization rates — didn’t explain the variation in joblessness nearly as well.

The idea that owning a home makes it harder to find a job because of higher moving costs is now known as “Oswald’s hypothesis.” And it’s come in for plenty of scrutiny. Some economists, for instance, have argued that this effect might be counterbalanced by the fact that people who own homes have denser local networks, which makes it easier for them to find jobs in their local area.

Now, however, Andrew Oswald and Dartmouth’s David G. Blanchflower have a brand new working paper (pdf) suggesting that homeownership has an even bigger and wider effect on unemployment than anyone has realized.