A couple years ago, New York Mayor Michael Bloomberg had a clever idea for dealing with cities like Detroit, whose populations are dwindling: let immigrants live there.

At the time, a spokesperson for Detroit Mayor Dave Bing dismissed the idea, saying, "There are things that would have to be considered before we just open the flood gates and say everyone come in." But new research from Jacob Vigdor, an economist at Duke, suggests that opening the floodgates a little bit might be worth it.

Vigdor's latest work, conducted by the Americas Society/Council of the Americas and Partnership for a New American Economy — the same pro-immigration group that recently released an excellent paper on North Carolina farm workers last month — looks, county by county, at immigration rates and housing values to see if there's any relationship between the two.

There is, and it's positive. Vigdor estimates that the average immigrant adds 11.5 cents to the value of the average home in his county. Considering that there are 40 million immigrants in the U.S., and 800,000 housing units, that adds up to about $3.7 trillion in increased housing value.

That could be a blessing or a curse. On the one hand, that's $3.7 trillion more wealth that somebody holds. On the other, it translates into costlier mortgages which in turn translate into costlier rents, making housing less affordable. But Vigdor's not particularly worried about that. In already costly counties like San Francisco or Manhattan, the effects are muted. "You don't see an effect at all in Manhattan or San Francisco," Vigdor says. "You see it in other neighborhoods that have fallen out of favor. If you look at what part of New York has had a big impact from immigration, it's the Bronx, it's Queens."

Areas like those, which saw middle class families flee in the 1970s, tend to develop more once immigrants arrive, not least because there's just more people around. "In a service economy, the number of jobs depends on the number of people to provide services to," Vigdor says. But immigrants don't just shift housing demand to blighted areas of metro regions that are wealthy overall, like the Bay Area or New York City. They also help out entire metro regions that otherwise would have undergone decline, especially in the Rust Belt. "The number of U.S.-born Americans residing in Chicago and surrounding Cook County, IL, has declined by 900,000 since 1970," Vigdor writes. "The arrival of nearly 600,000 immigrants over the same time period offset most of that decline—and most likely kept additional natives from leaving—blunting what could have been a catastrophic impact on the local housing market."

You can see this clearly in a map that Vigdor and his team developed to summarize his findings:

Vigdor relies on Census data, so as with any non-randomized study, it's important to be careful in sussing out causality. Maybe, for example, immigrants are just likely to flock to areas with high housing prices because they're likely doing better economically, and aren't actually causing the housing increase at all. To control for that, Vigdor exploited the fact that immigrants are likely to flock to cities where a lot of their co-nationals live, regardless of their economic prospects there. For example, there's a disproportionate number of Hmong immigrants in Minnesota and a large number of Vietnamese immigrants in Louisiana. Because it affects where people immigrate and not anything else that would in turn affect housing prices, the number of co-nationals living in a county is what's called an "instrumental variable" and can be used to isolate the effects of immigration more cleanly.

Effects of $1,000 or so on housing prices aren't going to make or break a family's finances, but Vigdor's work is a good reminder that immigration has sundry economic effects that aren't captured in simple analyses like the CBO's. Especially if the housing appreciation he finds is a result of immigrant spurring consumer demand, attracting new businesses to their neighborhoods and generally making those neighborhoods more attractive places to live, we could be talking about huge upsides here.

Update: I should add that Vigdor is hardly the first person to find this. Albert Saiz, then at the University of Pennsylvania and now at MIT, has found that an immigrant inflow equal to 1 percent of a city's population is associated with a 1 percent increase in rents and housing prices in that city. Thanks to Adam Ozimek for the pointer.