Beautiful Visualization

Chapter 8. Visualizing the U.S. Senate Social Graph (1991–2009)

Andrew Odewahn

IN EARLY 2009, MANY NEWS STORIES emphasized the collapse of bipartisanship. Although much of this reporting was of the typical "he said, she said" variety, one article in particular caught my attention. Chris Wilson, an associate editor at Slate, wrote a great piece in which he used voting affinity data and graph visualization to help explain Senator Arlen Specter's party switch (Wilson 2009). The graph showed two large party clusters (Democrats in blue, Republicans in red), connected by a few tenuous threads of senators who consistently voted across party lines.^[37] One of these was Specter.

The piece got me thinking on several dimensions. First, it was really cool to see quantitative evidence making the case for what was an essentially qualitative story. With one glance, you could see that something interesting was happening with Specter that presaged his break with his party. It made me wonder if there was similar evidence about other stories in the news. For example, a lot of reporting fixated on various Senate coalitions—the "Gang of Fourteen," the "New England Moderates," and the "Southern Republicans"—and how they were aiding or thwarting some initiative or another.

Basic civics would have you believe that the Senate, unlike the House, was designed by the Founders to dampen coalitions like these. It's a simple body: there are 100 senators, two from each state, who stand for election every six years. ...