govtrack
Measure bipartisanship based on difference in ideology scores
The bipartisan bills stats measures which members of Congress join on bills with members of an opposing party, but the stats are misleading. The top two "bipartisan" Senators are also the two Republicans with the lowest ideology scores (i.e. the most liberal Republicans).
A better measure of bipartisanship would be to somehow calculate the difference in ideology scores between the sponsors, such that a bill sponsored by both Jim Inhofe and Richard Blumenthal would be considered much more "bipartisan" than a bill sponsored by both Mark Pryor and Richard Shelby.
https://www.govtrack.us/congress/members/report-cards/2013/house/cosponsored-other-party https://www.govtrack.us/congress/members/report-cards/2013/senate/cosponsored-other-party
(Anybody can calculate who cosponsored bills across the aisle; we should take advantage of the fact that we've calculated ideology scores for each member of Congress.)
I had to scratch my head on this a few times. I like the idea. But I don't think it will work. The ideology score is based on cosponsorship, so in a way it's already measuring the same phenomenon as the bipartisanship % but in a different way. Another way of saying it is that there's no new information to be gleaned from combining the two statistics since they're already just two sides of the same coin. I think.
That said, check out Mike Enzi (https://www.govtrack.us/congress/members/michael_enzi/300041/report-card/2013) who is computed as being very conservative but also gets Democrats on his bills.