Source: BETSY MASON, 11/20/2013, WIRED
There is a temptation with any kind of data that has a geographical aspect to display it on a map. While maps are by far the best way to convey many of these data, sometimes they are not. This is one of those times.
Even though data on migration between states would seem to cry out to be mapped, this circular visualization by independent data journalist Chris Walker (@cpwalker07) can convey a lot of information far more neatly than a map. Patterns leap out that might have been obscured on a single map, or required many maps to convey the same information (see images below).
“It’s useful to think beyond maps especially for cases where you want to show interconnectedness between regions, which is what I was trying to do,” Walker wrote in an email to WIRED.
[NOTE: The interactive map discussed below can be found on WIRED’s site by clicking the link here. I should note that it worked best for me using Google Chrome]
When I first saw Walker’s migration circle (which he built using D3.js), it looked like a jumble that was impossible to untangle, but that’s before I realized it was interactive. If you haven’t already, mouse over the graphic to display information from single states to see where people from that state moved to last year, and where the people who moved into that state came from. More than 7 million Americans moved within the country, so if you’re looking for a time sink, you’ll find it in this circle.
“I think we can learn a lot from migration patterns,” Walker wrote in an email. “In a way, migration flows are one of the oldest forms of crowdsourcing. They tell you which geographies the crowd deems to be low-opportunity, and which the crowd deems to be high-opportunity.”
I live in California, so I started there. I love this state, and it seems to me people are moving here all the time. But it turns out, more people are leaving (73,345 more). As you can see in the snapshot to the right, Californians don’t usually stray too far though, tending to migrate to other western states. Or Texas. A likely explanation for the outflux is that you could buy a mansion in some parts of the country for the same amount as a 2 bed, 1 bath house in the Bay Area.
Despite the initial jumble, a couple of things jump out before you even begin interacting with the graphic. A lot of people are moving out of New York (135,793 net loss) and ending up all over the place, including the other side of the country. The same is true for the Midwest, with people mostly landing in the Southeast, Southwest and California. These trends are clearer when you look at individual states, but the broader trends would be easier to grasp if you could mouse over the region names to see where, say, everyone in the Northeast moved.
One thing to note as you look at what’s happening with New York, for example, is that only exchanges of at least 10,000 people are depicted. This, Walker says in his blog post, is to keep the graphic from becoming to messy. So, while many people are moving into New York, most of them aren’t shown because they are coming from many different places in smaller numbers.
Some of the other things Walker noticed include that fact that a lot of people are moving to Florida, many of them likely retirees. “Interestingly the state contributing the most migrants to Florida is neighboring Georgia,” Walker wrote in his blog. “Texas, New York and North Carolina are the next largest contributors.”
The second largest draw for migrants was Texas. “Over 500,000 people moved to Texas in 2012,” Walker wrote. “People tend to come from the Southeast, Southwest and the West, with the biggest contributor being California. 62,702 Californians packed up and moved to the Lone Star state in 2012.”
People who leave D.C. don’t really leave, generally moving next door to Virginia or Maryland. In contrast, people moving from Maine and Alaska are chasing the sun all the way to California and Florida. Check out his blog post for more insight.
In contrast to Walker’s circular visualization, old census atlases used maps to show the migration data. In August, we visited the Prelinger Library here in San Francisco and took a look at some of these atlases in their collection. In the images below, you can see how the data was displayed. On the right is a map showing where New York natives lived in 1890. New York has the most natives, unsurprisingly, and is the darkest. But the migration pattern is similar to 2012 with New Yorkers heading all over the place. These maps highlight some of the limitations of maps for displaying this kind of data.
Maps from the 1890 Census Atlas at the Prelinger Library in San Francisco. (Ariel Zambelich/WIRED)