These ratings are from Knowlegis, and reflect the first half of the 110th Congress. The ratings and rankings data are here. Their criteria are described here. Yes, they have caveats.
I’ve rendered their Senate data as a graph. Higher scores on the x-axis indicate more power. Intuitive results abound:
* Most Democrats are towards the top, and most Republicans towards the bottom. Power depends in part on majority status.
* Party leaders and committee chairs are toward the top.
* Seniority matters. Many members at the bottom are freshmen.
The most subjective portion of their ratings is the “Sizzle/Fizzle” factor, which accounts for why Obama, Clinton, and McCain are towards the top — despite little institutional power in the traditional sense — and Larry Craig is at the bottom.
I have plotted the data to highlight the ratings and downplay the rankings. Rankings are often dubious measures because they distort the underlying ratings by placing observations on an interval scale. In terms of ratings, Carl Levin and Robert Byrd are .04 units apart (45.08 and 45.04, respectively) while Harry Reid and Ted Kennedy are about 45 units apart. In terms of rankings, each pair of Senators is one unit apart.
The plot reveals that a lot of Senators have roughly similar amounts of power, despite the rankings. Without any measure of uncertainty, however, there is no sense of how much of a difference is really meaningful.
And, of course, it’s not clear what these power ratings would explain that could not be captured by seniority and leadership positions.
[Hat tip to Scott Adler.]