Almost every day, you’re likely to see a map of the United States and its fifty states. These maps are so familiar and so common that, through force of repetition, they make it seem as though those state boundaries mark out the natural and inevitable division of the world into a cellular mosaic of local geographies. And it’s not just seeing those lines on a map which makes the boundaries seem so powerful. On your driver’s license, on your tax return, and on your ballot, the way we categorize ourselves into members of different geographic units has enormous consequences.
But do the lines of those state boundaries really match the actual unit landscapes in which people go about their lives? For instance, doesn’t it seem like Kenosha, Wisconsin, might have more similarities with the north side of Chicago than it does with Superior, Wisconsin, in the other corner of the state? The boundaries of states were drawn for all kinds of idiosyncratic historical reasons. Pennsylvania and Delaware are separated by the sweep of a 12-mile circle traced around the colonial town of New Castle. Michigan got the Upper Peninsula as a consolation prize for relinquishing its claim on Toledo. And even borders which came into being for less capricious reasons may have been rendered obsolete by new technologies, migration patterns, or economic pressures.
Geographers have been trying to discover what are sometimes called functional or natural regions for more than a century, often with the goal of criticizing and reforming the geographic divisions drawn by politicians. After all, doesn’t it stand to reason that ancient boundaries may not make much sense in the modern world? As H. G. Wells complained in 1902 to the Fabian Society, “The areas in which we shape our public activities at present, derive…from the needs and conditions of a past order of things. They…still preserve the essential conceptions of a vanished organization.”
But there’s a catch to this rationalistic way of thinking. If we want to make an empirical argument about the geography of new functional regions, what types of evidence should we use? Carl Sauer thought the pattern of the cultural landscape could be used to delineate “region[s] of geographic unity.” Bill Cronon used data about credit flows in Nature’s Metropolis to argue that the vast majority of the Midwest was functionally integrated into the city of Chicago, effectively merging “city” and “hinterland” into a single economic unit. And there are many other types of variables we could use, too—marriage patterns, linguistic variation, ecosystem energy flows, and so on.
While there’s no one perfect type of evidence for making these kinds of regionalization arguments, the availability of new kinds of big data does make it possible to explore geographic connections in interesting ways. My colleague Alasdair Rae and I published a paper in the journal PLoS One last fall where we turned to a variable which offers a good proxy for both the spatial patterns of how people live and how areas are economically connected: commuting data. Using a data set from the US Census Bureau, we mapped more than 4 million commuter connections between communities in the United States. We then subjected this data to a community-detection algorithm. That’s a high-powered piece of computer software that groups data points together in a way which keeps highly-related points in the same category and splits weakly-related points into different categories.
What we found in this analysis was that it’s possible to detect a geography of “megaregions” in the United States, where each megaregion is an interrelated constellation of large, medium, and small communities tied together by a dense weavework of commuter connections. Here’s what that looks like on a zoomed out scale:
We turned the results of the paper into an interactive web map which lets you explore the pattern of commutes and the borders of the new megaregions in more detail. For instance, here’s Madison; you can see the radial communities of Dane County as well as the tug of commuters eastwards towards Milwaukee.
It’s important to realize that the borders we’ve drawn between regions don’t indicate that nobody is crossing them. To the contrary, some neighboring regions have a lot of interconnections, like the “El Asfalto” (Los Angeles) and “Portola” (San Diego) regions:
Others are more isolated, like the “Blue Ridge:”
There are many places where we can see that state borders really are just invisible lines. Michigan, for instance, gets to claim Toledo back:
But there are also places where it’s possible to see state borders retaining their prominence, like on the eastern and western side of Connecticut. That tells us something interesting about how state lines might be affecting commuter patterns—perhaps people are marginally less willing to work in one state and live in another.
Looking at a map like this at first feels unfamiliar—where are all the shapes of the states we so instantly recognize? But exploring the place you live, you might find that these new regional delineations might feel more familiar than old state lines. For instance, one reader who is from Roanoke, Virginia spoke of always feeling distant from the state capitol in Richmond, and more connected with the culture of North Carolina. Well, it turns out that, from the point of view of commuter interconnections, Roanoke really does become part of a North Carolina corridor stretching out to the ocean.
So what do we do with these new megaregions? It’s pretty unlikely, given the federal structure of the United States and the deeply entrenched power of state-by-state institutions, that we’ll see any political boundaries get redrawn anytime soon. But it’s still possible for policy action to recognize these new regions as the kinds of geographic units which are appropriate to manage and plan as a whole. For instance, as river management became a national policy priority at the beginning of the twentieth century, the federal government encouraged interstate cooperation to deal with the many places where watershed boundaries cut across state lines. Congress even created an entirely new agency, the Tennessee Valley Authority, in order to make the geography of state power match the geography of an ecological and economic problem. Today, when building new transit systems or selecting sites for expanding the housing stock, it would make sense to start with the commuter megaregions in mind.
If these new geographies only take hold in the minds of experts and technocrats, however, they’ll never get very far. Far more powerful is the kind of organizing that becomes possible when people begin to think of each other as neighbors across new territories—thus including new groups of people into an expanded body politic. If everyday people can imagine themselves as “belonging” to new places like Corntassel or Winnebago or Yanaguana, then the foundation stone for a broader and more substantial cooperation will have been laid.
Finally, it’s worth pointing out one kind of line which this map doesn’t do anything to critique: the national boundary of the United States itself. In this case, that’s because we’re drawing on data from the Census Bureau, and that data set doesn’t include people who commute outside of the country. Even if we did have data about international flows, the hard lines which are drawn around the nation-state make it exceedingly difficult to create new geographies that cut across national sovereignties (as the retrenchment from regional cooperation to national jingoism is currently illustrating in Europe). But even if national borders seem to be drawn in increasingly thick cartographic lines, we should still remember that they are nothing more than historical accidents. The lines drawn by people in the past can be redrawn by those of us in the present—if we are creative and willing enough to try.
Featured image: screenshot of “The Megaregions of the US” web map by Garrett Dash Nelson and Alasdair Rae highlighting the megaregion of Winnebago.
Garrett Dash Nelson is a historical geographer at Dartmouth College in the Department of Geography and Society of Fellows who studies the history of landscape, planning, and community in the United States. He received his PhD from the University of Wisconsin–Madison in 2016 and was a member of CHE from 2011 to 2016. Website. Twitter. Email.