Regular readers of ecopolitology know I have a thing for maps. But add a dynamic component to a map that depicts political change on a spatial and temporal scale not bound by political borders and the resulting map looks more like the temperature or precipitation forecast maps one might see on The Weather Channel than any electoral map. And that's exactly what David B. Sparks, a PhD candidate in political science at Duke did with presidential election data from every election between 1920 and 2008. (See video below)

As opposed to a "simple" choropleth map which depicts data constrained by political units like counties or states, an isarithmic map can capture the existence of regional patterns otherwise obscured by the superimposition of geographically arbitrary political boundaries.

As Sparks explains on his blog, "Isarithmic maps are essentially topographic or contour maps, wherein a third variable is represented in two dimensions by color, or by contour lines, indicating gradations." In this case, the third variable is partisan intensity, represented by shading.

Unconstrained by county and state political boundaries, the animation, writes Sparks, "does a good job of depicting local 'peaks' and 'valleys' of partisan support clustered around urban areas."

The animated interpretation also highlights macro dynamics in American politics including the widespread and lengthy support for President Roosevelt during the New Deal of the 1930s and 1940s; the shift from a Democratic to a Republican-dominated South; to the familiar coastal-interior division characteristic of today.

While there are some methodological troublespots in using isarithmic maps to depict political data, I applaud Sparks for his creativity and look forward to revisiting the map in 2012 — even if the actual content of the map turns out to be difficult to stomach.