|View From a Height
Commentary from the Mile High City
Wednesday, June 16, 2004
Is much more than what we do know. With Jon and Ben going back and forth over the Club for Growth's Republican Senate poll and Colorado election history, we need to back off and note some limitations here.
First, I agree with Ben that I can't see any obvious problems with the poll's methodology. They pulled 500 likely Republican primary voters. Now the difference between the two is well within the margin of error, 6 weeks away from the primary itself. This is a little like the difference between being 30-32 in June, and being 30-33. I know which one I'd rather be, but I wouldn't lose sleep over the difference.
However, the CFG probably backs Schaffer. (Coors won't be showing up on any RINO lists, though, since those are for voting representatives, not candidates.) The real problems come in the refining questions, and therefore the analysis. The accuracy of a poll is roughly proportional to the square of the number of participants. With only 200 most-likely-voters, Schaffer's 10-point lead is still well within that sub-sample's margin of error. It's likely that fewer than 200 voters came from Schaffer's old district, rendering any data from that subsample almost completely meaningless. At these numbers, any discussion reminds me more of Isaac Asimov's "Election Day 2084" than of serious political analysis. It doesn't speak well of the Club that they attempt to draw conclusions from such small samples.
Ben's reminds us of Bruce Benson's campaign. While 1994 was certainly a Republican year, Roy Romer was a popular governor running for re-election during an economic boom. At the time of voting, the state had a Democratic governor, and a Democratic Senator, Ben Campbell. It barely went for Owens in 1998, and barely went for Dole in 1996. It was a much different place than now.
Bruce Benson was a state party chairman, who fits as much into the Bill Armstrong mold as the Pete Coors mold. As does my recent professor Buie Seawell, state Democratic party chairman who ran for Senate in 1990, and couldn't even get the nomination.
In short, I don't think this poll, or the state's history from a decade ago, have much to tell us about what will happen in November.