So, a lot of political maps are circulating these days. My Washington Post colleague Reid Wilson has noted one map, by journalist Colin Woodard, that divides the United States into 11 “nations.” Another, by journalist Dante Chinni and political scientist James Gimpel — which we featured here — creates a more variegated patchwork. My Post colleague Chris Cillizza, comparing districts held by Republican and Democratic members of Congress, writes:
The simple fact is that there is almost no overlap — culturally, politically, economically, socially and virtually every other word ending in “ly” — between the districts represented by Republicans in Congress and those held by Democrats. If the two sides often seem like they are talking to two totally different electorates on every issue, it’s because they are.
Of course, it’s true that Americans aren’t of one mind on many political issues. But it is important that we not look at these maps and infer that we are so politically polarized by geography. In fact, most Americans live in places that are at least somewhat politically and ideologically diverse — even if that’s not reflected in how congressional district boundaries are drawn. In terms of the most important driver of political choices — partisanship — most of us live in a purple America, not a red or blue America.
Here is a map of the 2012 presidential vote at the county level by the University of Michigan’s Mark Newman:
As Nate Cohn rightly points out, this kind of map makes the country look more “red” than it really is. A lot of those solid red counties have small populations. Also, this map only captures the party that won each county, not the county’s underlying partisan composition — which often reflects a mix of Democrats and Republicans. Once you take into account population and this more purple mix, you get a U.S. map that looks like this:
The sparsely populated red counties shrink to thin lines. There are a few patches of dark blue in some big cities. But mostly the map is colored with various shades of purple. The same was true in 2004 and 2008.
There are other maps that make this same point. This one by Chris Howard, for example, uses shades of red, blue and purple, but captures population by lightening and darkening the hues. The washed-out areas of the map are where fairly few people live. You can see how much this improves over the three maps above it.
All of this is consistent with how America looks when you focus the microscope even narrower — to voting precincts within countries. As I noted the other day, most precincts are not solidly red or blue:
(Graph by Eitan Hersh)
This is also consistent with political science research showing that red and blue states are not that polarized. Political scientists Matthew Levendusky and Jeremy Pope constructed measures of economic and social liberalism and compared them in red and blue states based on the 2004 presidential election. There is tremendous overlap:
Graph by Matthew Levendusky and Jeremy Pope.
Levendusky and Pope go on to show that the differences between Democrats and Republicans within any individual state are more pronounced. Partisans are much better sorted ideologically these days, but this doesn’t have much to do with geography.
Of course, there are interesting differences among American communities — economically, demographically and politically. Provo is not Berkeley, and vice versa. The point is that most Americans don’t live in Provo or Berkeley. Most of us live in places that are, politically speaking, somewhere in between.