Yesterday the Senate defeated Tom Coburn’s amendment to kill the National Science Foundation’s political science program — which had my colleagues up in arms — by a 62-36 vote that broke down largely along party lines:

To understand the vote breakdown better, we can use this graphic, which was automatically generated by Royce Carroll using the Lewis-Poole Optimal Classification estimates for the 111th Senate:

Each senator is placed at their estimated ideal point in the ideological space. The diagonal cutting line, which represents the best-fitting line dividing yes from no votes in the space, indicates that the vote reflected both the primary ideological division between the parties (in this case, cutting “wasteful” government spending) and the second “social issues” dimension (feelings toward pointy-headed academics?).

For a different perspective, here are regional breakdowns of the vote from Govtrack.us with both a standard US map and one that is deformed to give each state equal area:

Note to Senator Coburn: This post was not funded by the NSF.

Update 11/9 7:02 AM: Stanford’s Simon Jackman posted this graphic portraying the relationship between estimated ideal points along the primary ideological dimension* and the Coburn vote (click the graphic for a larger version):

It’s especially useful for distinguishing the senators whose votes are not well-explained by their estimated ideal points in the first dimension.

* Jackman’s estimates come from the Clinton-Jackman-Rivers algorithm for scaling roll call votes, which differs slightly from the Lewis-Poole algorithm used to produce the circular graphic above.