Election Graphics
I love maps, and so have been fascinated with all the color coded maps that were used during Tuesday’s election coverage in the media. Boing Boing just posted a map by Jeff Culver that represented the states not as the polarized completely red and completely blue, but instead more of a spectrum between the two, since most of the states do not vote 100% Democrat or Republican. So you definitely get a different picture looking at his map. What I see is that the Eastern half of the country is a fairly similar mix, going from mostly bluish purple in the north to a more reddish purple in the south, with only Indiana and Alabama standing out from their neighbors in red intensity. The western half of the country seems more bit more polarized in terms of states, with the western and southwestern states and Colorado tending towards the blue while the plain states, Montana and Texas tending more red, and finally Wyoming, Idaho Nebraska, and Utah standing out as beacons of highly saturated red.
Of course this graphic, as most, doesn’t tell the whole story. It’s just one way of looking at the demographics. Part of the problem is that we are seeing things on a state-wide level here. What I think might be more useful would be to take the map and section it out by county. Even better would be to do it by precinct, but then if you are looking at a nation as a whole, precincts might be too small to make out on such a graphic. What one really needs is a way to smoothly zoom in and out, with colors changing as smaller areas of detail become discernable. Of course that is a monumental task and ultimately the data simply wouldn’t be there to zoom very far in, since there is to my knowledge no data that goes lower than precinct level. But theoretically you could get to street level or at least a 5-square-blog area. Getting to individual homes would be impossible both for reasons of privacy and because many people live in single households. Then you encounter another variable, which is population density. None of the current graphics represent this important factor, but I suppose you could do this by creating a third dimension and adding elevation to more densely populated areas.
In any case, even before I saw this I was envisioning a graphic that would represent the entire country or at least one state in a more “granulated” but softer way. So Jeff Culver’s map motivated me to take an image of Virginia that I got from CNN’s election site with a breakdown of counties. CNN uses lightness to show how close a vote is, so a completely white county means the vote was almost exactly 50/50. The darker or more saturated the color (whether red or blue), the more lop-sided the vote was to one candidate or the other. I just plopped this into Photoshop, got the harsh county delineations out, and filtered it a bit to make it more blurry. A collegue also informed me that the colors on the graphic aren’t exactly red and blue, but more orange and purple. Being color blind this wasn’t obvious to me, but I didn’t make any specific changes to the color from the original except to make it slightly more saturated…
I even considered doing this for every state, but sitting here doing this for 12 hours didn’t seem worth it. Maybe I can find a similar graphic of the entire country and just do the same thing for that one…