Greetings from the People's Republic!

Recently both Calpundit and Dan Drezner took up the question of how "extreme" Bush and Kerry are (in the sense of "How far to the right is George Bush? How far to the left is John Kerry?") There were following on an article in the Chicago Tribune by Political Scientist Jeffrey Jenkins.

Jenkins uses what I consider to be some fairly impressive algorithms, developed by Profs. Keith Poole and Howard Rosenthal, to rank order politicians on how "conservative" or "liberal" they are. [Note that I use quotation marks in the previous sentence not because I'm some sort of "air quote" freak, but because in our modern political climate the classic meanings of liberal and conservative have been completely lost. Witness our "conservative" president who runs massive deficits, increases government spending almost rampantly, and favors both amendments to the constitution and "public works" programs to advocate certain types of personal behavior over others.]

There has been an excellent blogosphere discussion of this subject, and I recommend reading through the comments on Dan's and Kevin's sites, particularly the follow-up posts by Dr. Jenkins. For background information, Keith Poole maintains an informative web site which describes the underlying methodology. Looking over that site I was struck by five things, which I'll list here:

1. The Poole-Rosenthal method is an example of political scientists doing actual quantitative analysis, as opposed to just spouting about their feelings or gut instincts. Whatever else you think of their results, you have to respect them for making a serious, measurable effort.
2. Poole is the Kenneth Lay Professor of Political Science at the University of Houston. Sorry, I just can't let that pass without a smile. He received this honor in year 2000, when it must have meant something slightly different from what it means today.
3. Poole's method (described in great detail on his site) ranks politicians on various axes --- kind of a political Meyers-Briggs test --- by performing an iterative ordering based on their public votes until the most self-consistent result is achieved. Basically, he removes the subjective questions of "is this particular vote liberal or conservative" and instead orders politicians relatively by how often they agree in their voting patterns. This requires an analysis of an enormous number of configurations, so the Poole-Rosenthal method is yet another exmaple of something useful made possible only through the use of modern computer technology.
4. Looking through Poole's liberal/conservative rankings of our current Senate and House, the degree of polarization compared to party affiliation is stunning! In the Senate, there is only one outlier (not counting Jeffords, a Democrat in Independent's clothing). Zell Miller (D-GA) is noticeably more conservative than any of his Democratic brethren. That's it. Otherwise, the Democrats (plus Jeffords) and the Republicans have a pretty clean break (and Democrats should be trying to pull over, DUH!, Snowe, Collins and Chafee).
   
   In the house, it's even more extreme. Not a single Democrat ranks as more conservative than a single Republican. In part, this makes me wonder whether Poole is measuring "ideology" or "party loyalty." He has a lot of historic information on his site, and I hope to post more later once I have a better understanding. Either way, the partisan divide is shocking, probably the worst since the pre-Civil War era.
5. Looking at Poole's results for the 108th House of Representatives, who should be at the top of his list (ordered from most liberal to most conservative)? That is, who, according to Poole, is the farthest out left winger in the entire congress? None other than my own Congressman, Jim McDermott.

Greetings to all from the People's Republic of Seattle! Further messages will be sent once we get vaguely in contact with reality!