The Dem senators broke between a hard count of 20-odd who were prepared to vote against cloture, win or lose; diehard ayes; and the rest who were (my surmise) prepared to be the one to put the anti-cloture vote over the top, but not be a foot-soldier in a losing battle!

When it was clear that cloture would pass, this last group were ecstatic to have an 'alibi' for goosing their constituents with an aye.

But how important is constituents' ideology anyway?

I looked at the DW-NOMINATE 1st dimension scores for the Senate in the 108th Congress.

For states with two Dem or two GOP senators, the difference in ideology between them is relatively small: the mean and standard deviation of Dem scores is 0.112 and 0.110, of GOP scores 0.120 and 0.108.

But for states with one Dem and one GOP senator, the numbers are 0.626 and 0.177.

In the 108th, the average Dem and GOP scores were -0.385
and 0.369, a spread of 0.754.

Thus the average spread between senators representing a single state is almost as wide as that between the two parties as a whole.

What sort of ideological coherence is that?!

For example, in Illinois, both Durbin and Fitzgerald were first elected to the Senate in the late 1990s. Yet, in the 108th, their scores were -0.538 and 0.260, a spread of 0.798!

Clearly, one's not dealing here with either an incumbency effect or the sort of historical loyalty that slowed down the Dem's loss of the South in the decades following 1964.

Now, DW-NOMINATE isn't perfect - it doesn't take salience into account, so far as I know.

But surely these results aren't just an artefact?