MIKE'S MATH

Michael Tomasky chooses a very strange approach to claim some quantitative heft for his otherwise well-stated case that the Democratic Senate Caucus will continue to represent a range of views whether it includes Joe Lieberman or not:

You start with their National Journal numbers -- specifically, their liberal support score for 2005. This score is defined in this way: If Senator X has a liberal support score of 90, it means she is more liberal than 90 percent of her Senate colleagues..So, off the top of your head: How many of the 44 Democratic senators have a 90 or better? Nine? Ten? Try four...

If this sounds like a meaningful measure of how liberal Senate Democrats are, or how broad the range of ideologies among Senate Dems are, then go back and read that second sentence again. According to Tomasky's description, the National Journal rating (yes, that's the same one that gave us that silly talking point about Kerry and Edwards being the 1st and 4th most liberal senators) is a stanine (remember standardized tests?). It measures how liberal a given senator is as compared to the other 99 senators (the system must be more complicated than Tomasky's describing it, because it's physically impossible for Ted Kennedy to be more liberal than exactly 96.7 other Senators). Which (lest our friends at the National Journal take offense) may be useful to know in evaluating a particular Senator, or even a few of them. But in terms of looking at a 44-member caucus, it's less useful. It could tell us (assuming we accept the rubric for the calculations, which Tomasky goes on to say he doesn't) whether there's any overlap along the scale between the two caucuses - that is, whether Lincoln Chafee is more or less liberal than Ben Nelson. It could even tell us something about how the senators are spaced along the ideological spectrum they represent. But knowing that the Democrats have four Senators in the 90s and "a passel of B's", while the Republicans have

have just three 90's: Jeff Sessions, Wayne Allard, and Tom Coburn. But they do have more in the 80's

sheds precious little light on the question Tomasky is trying to answer: How ideologically diverse is the Democratic caucus (rather than how the Democratic Senators are spaced along the ideological territory of the caucus). Maybe there's an argument to be made about how the ideological breadth of one caucus skews the distribution of the other caucus along the spectrum of all 100 senators, but I don't think Tomasky is making it. His argument seems to be that if the Senate Democratic Caucus were really full of Ted Kennedys, you'd see more of its members scoring in the 90s. But if the Caucus were full of Ted Kennedys, it would become that much harder for Ted Kennedy to eke out a 90. Because, as they say, it's all relative. If you took a snapshot of the current distribution of Senators along the National Journal scale, on the other hand, you would have a tough time (unless you were, say, Jacob Hacker) telling from looking at it whether you were looking at the Senate circa 2006, 1936, or 1846 - because changes in the ideological breadth of the Senate would only translate indirectly into changes in the spacing of the Senators along that breadth. And you'd be no closer to figuring out how the ideologies represented by the folks in the Senate compare to the breakdown of America, or even Connecticut. That is, if I understand the National Journal ratings correctly. If I'm confused, then forget it. If not, then Tomasky's parallel universe of Democrats who all score in the 90's bares a strong resemblance to Garrison Keillor's apocryphal town in which "all of the children are above average."