Lance Fortnow, a scholar of computational complexity, has put together a map integrating Tradesports data on Senatorial and Gubernatorial races. You can see it at http://www.fortnow.com/governor/ and http://www.fortnow.com/senate/. The predicted governors are most interesting to me. This doesn’t look like a Red State/Blue State map. This has Republicans in California, Vermont, and Connecticut. There are Democrats across the Mid-West, even down into Tennessee and Arkansas.

If the parties are as all dominating as the media represents, and the nation as polarized as we’ve been told, this shouldn’t happen. This map, often showing different parties winning the local and national races, shows a healthy polity.

The Senate map is less excellent, but still far from bad. Only Missouri has a tightly contested race. On the other hand, it’s not like these are percentages of voters: they’re colored based on the perceived probability of an outcome. An absolute certainty of a 51/49 election will be bright blue or red.

Where’s all this healthy political intermingling coming from? Why is sushi-eating Times-reading Connecticut electing a Republican? Why is Oklahoma electing a Democrat? I’m hesitant to read too much from one complicated datum. But perhaps the problem isn’t the parties, or the Bush-hating ANSWER minority, or the crypto-fascist anything-Bush-says minority. Maybe it’s Washington, D.C. Maybe we can have less-polarized politics out in the states, but not when three hundred million people are being represented by less than six hundred.

(It is also worth noting that none of these races are gerrymandered, their borders having been drawn when America had something better to do than squabble in the sandbox.)