These are a set of diagrams taken from a recent WSJ article talking about how, "the pandemic changed where Americans live." I know that this is a topic that gets a lot of air time (both here on the blog and elsewhere), but these diagrams do a good job of showing the flow of people, as well as how things may have changed/accelerated since 2018.

These diagrams also remind me of the work of Charles Joseph Minard. A French civil engineer, Minard is best known for his contributions to the field of information graphics, and in particular his flow maps. His most famous piece of work -- which I happen to have hanging at home -- is his depiction of Napoleon's losses during the Russian campaign of 1812.

The map itself is from 1869 and is packed full of information. It shows the number of Napoleonic troops as they left for Moscow, the distance they traveled, the outside temperature (the French weren't properly prepared for the cold), latitude and longitude, the direction of travel, and the location of the troops relative to specific dates.

The point of the diagram was really to show how disastrous this campaign was for Napoleon. The thick beige band on the left is showing over 400,000 troops setting out. But by the time they reached Moscow -- which, by the way, had been abandoned before their arrival -- only about 100,000 troops were left.

The thin black bar on the bottom is showing how many troops ultimately remained and returned at the end of the campaign -- the number was only about 10,000. So the vast majority of Napoleon's troops perished. Supposedly over half either starved or froze to death.

Some 150 years later, and we are still using flow charts to clearly depict the movement of people and things.

For years, the data has been clear. Many Americans are moving from expensive cities, like Los Angeles, to less expensive metropolitan areas like Dallas-Fort Worth.

But Wendell Cox's recent article over at New Geography is a good reminder that these data sets can be limited. The US Census Bureau currently tracks domestic migration at the county level only. This can be a bit of a problem as counties vary dramatically in terms of geography and population.

The New York metropolitan area, for example, is comprised of 25 different counties averaging about 750,000 residents. The Los Angeles metropolitan area, on the other hand, is compromised of two counties averaging about 6.6 million residents.

These sorts of nuances become important when you're trying to figure out things like whether people are moving to/from urban cores or the suburbs. Case in point: The San Diego metro area is compromised of a single county. When people move there, the data says nothing about how urban or suburban they might be.

Dallas-Fort Worth is a lot easier to read. Since 2010, it has had the largest net domestic migration of any metro area in the US: +443,000 residents. But county data reveals that it is entirely suburban. The core (Dallas County) actually lost 57,000 people from 2010 to 2019. And this is not unique to the Dallas-Fort Worth area.

Photo by Gabriel Tovar on Unsplash

In response to President Trump’s proposed immigration bill, Brookings recently analyzed census data from earlier this year to demonstrate the importance of immigration for growth within much of the United States. 

I’d like to share three tables from their analysis.

The first two look at international migration grains and domestic migration gains over the last 3 decades (the last decade isn’t quite a decade).

Here you can see that New York, Los Angeles, and Miami (all port cities) have dominated international migration to the US since 1990. But at the same time, international migration has become less geographically concentrated. From 1990-2000 the top 5 cities received almost half of all immigrants moving to the US. More recently, that number has dropped to 34%.

Domestic migration is different in that it’s a zero sum game. When one US city gains, another US city loses. Here there is a very clear migration trend toward cities in the southwest – arguably because of weather, job growth, cheaper housing, and probably a bunch of other factors.

If we look at actual international and domestic migration numbers over the last 6 years, the 12 largest metropolitan areas look like this:

The key takeaways here are that 8 of these cities are losing people to domestic migration and only 7 of these cities have a positive net migration number – meaning their population is actually growing.

What is clear is that the international migration column is a pretty important one if you believe that growth is valuable. 

If you’re Dallas, Houston or Atlanta, maybe you care a little less about that column. But for most of the other cities, international migration is either the only way you’re growing (look at Miami go) or keeping your population losses in check (see Philadelphia).