Since the last election, you’ve seen all stripe of maps purporting to show us what really went on with the presidential race. Campaign spending-cartograms that showed the continental United States as a distended blob; maps that rendered votes as pointillist dots. But this one, created by Princeton Professor Robert Vanderbei, takes a slightly different approach. It’s one of the few that shows the action in three dimensions.
Like some other maps we’ve seen, Vanderbei’s displays votes county by county, eschewing solid reds and blues for a palette of purples, graded according to how each county voted. But what makes Vanderbei’s visualization unique is his use of the z-axis to show how many people voted in these counties. In metropolitan areas, columns shoot up like neon skyscrapers; in flyover country, it’s typically more of a low-rise affair. But the effect is powerful: At a glance, Vanderbei’s map shows not just how the country voted, but where it voted, too.
And that means cities. The democratic lean of places like New York City, Los Angeles, Philadelphia, and Boston shouldn’t be news to anyone, but seeing the results like this gives you a sense of just how overwhelming the number of voters really is in those densely populated urban centers. And so long as those towers stay blue, it’s why you won’t hear about states like California or New York being in play anytime soon.
Vanderbei, a professor of operations research and financial engineering, made his first "Purple Area" visualization after the 2000 election. He had been reading USA Today when one of the typical "county-by-county, red-blue" graphics caught his eye. That map, he says, "made me wonder why anyone would paint a county-by-county map in such a way as to imply that a county has cast its vote for one candidate or the other. I live in a county that went about 52% republican and 48% democratic in that election. Painting the county red seemed highly misleading."
Vanderbei says he’s played with the idea of a 2-D graphic that accounts for population density or voter turnout, something where these variables are reflected by the intensity of the color, but ultimately the discrepancy between cities and rural areas is so great that you wouldn’t end up with much of a map at all. "The trouble with that approach," he explains, "is that the difference between densely populated places, like Manhattan, and sparsely populated places, like counties in Montana, is so great that a linear intensity scaling renders almost the entire U.S. as virtually black, with just a few of the highly populous counties showing up." And even if he did take that route, the final product wouldn’t be nearly as stunning as this one, available on the Princeton site as an interactive, 50 MB WebGL file. Good visualizations need not be beautiful, but the sci-fi chic look of this 3-D visual certainly doesn’t hurt its chances of grabbing your attention.
Of course, as long as we continue to operate under the Electoral College, state totals are really what matter in the end, and you’d be hard-pressed saying who won after a quick look at this map. But a more localized breakdown is always insightful, and Vanderbei says he can "understand that people would like to get a sense of how things are in their 'neighborhood,' as that might be different from the statewide result."
But my biggest takeaway was that the Obama campaign’s ground game in Colorado must’ve really worked. Denver’s the country’s 23rd most populous city--smaller in population than Houston, Dallas, El Paso, and Fort Worth to the south in Texas--but its impressively tall (and solidly blue) stack represents a key source of votes in what was a hotly contested state.
[Hat tip: Technology Review]