I tend to avoid Political commentary on this blog, save for issues concerning the economy.  However, this morning I stumbled on some data — or more specifically, data presentation — which will be of interest whether you are a republican, democrat, or something else entirely.

Robert Allison, writing on the SAS Learning Post this morning, has a great piece titled, “Building a Better Election Map“.  Allison notes that we are all confronted with a congressional election map that looks something like this:

This is misleading on a lot of levels.  From a republican perspective, it implies that they still have control of the vast expanse of America.  For democrats, this map makes them question whether or not they really took control of the House of Representatives.  It’s simply not a good way to combine the population distribution of the U.S. with the data on House representation, which is supposed to be apportioned according to that population distribution.  Allison experimented with a number of formats, and ended up with a great interactive map that divides the U.S. up into 435 equal sized representational images and then color codes them according to the current representation.  Note that this map also shows where “flipped” seats happened this year.

Well, ain’t THAT neat!  This immediately lets the reader see where geographic trends are happening.  Several interesting pieces of data come out instantly.  For one, Texas is “bluer” than one might think.  Second, the largest number of republican-to-democrat flips happened in pivotal Pennsylvania (not in California, where one might have thought listening to the newscasts).  Third, the old Confederacy is a lot “bluer” than one might have thought, with democrat-to-republican flips happening in Virginia (2), South Carolina, Florida (2), Georgia, and Texas (2 thusfar).

As noted, Mr. Allison’s work is interactive, and I highly recommend you read his entire piece.  It’s a great article on both politics as well as data representation.