Speculations on the upcoming elections, using the RM’s skeptical Maxims about Polls and Democrats

Ok, we’re about up to the 2006 midterm elections, and all the pollsters except for Karl Rove are predicting a Democratic sweep. Most are predicting the Dems will take over one House of
Congress. Some are predicting Dems to take over both Houses. I disagree. [UPDATE 11-8-06 - I was wrong of course. Dems take House, and the Senate is still up for grabs as of this writing, because of a couple of too-close-to-call-races.]

For some info about Polls Polls Polls, check out Pollster.com. As you read the polls and the interesting analysis on the Pollster.com blog, keep in mind my own skeptical maxims.

The RM’s Skeptical Maxims about Polls - The Maxims are based on the apparent fact that Dems don’t match what they tell pollsters, but Repubs do.

I hope I’m wrong! I’d love to see the Dems take control of both Houses of Congress. But I’m very skeptical. Read on, and see where my skepticism comes from. I’m in agreement with Karl Rove, that the Repubs will keep both Houses. And here’s why ———————–

• Voters tend to overstate (to pollsters) their willingness to vote against the incumbent. When in the voting booth, a lot of people who told pollsters they’d vote against the incumbent change their minds, because the known is safer than the unknown.
• If the incumbent is a Republican, the Dem challenger will lose unless he is polling consistently about 5% higher than the incumbent. This is because Republicans are better at getting themselves motivated to come out to vote. Dems tend to be lazier about getting to the polls. Dems tell pollsters they are likely voters, but then stay home. We’ve seen this phenomenon repeatedly for generations. We just saw it in a big way in 2004.
• Likewise, If a seat is open, the Republican will win unless the Dem consistently polls substantially higher, perhaps 5% or more higher. If the Dem polls roughly equal, up to slightly higher, the Republican will win.
• If the incumbent is a Democrat, the Republican will unseat him if the Republican consistently polls about equal. That’s because Republican turnout will come closer to matching poll answers than the Dems will. The Dem will get less than the percentage polled, the Republican will come closer to hitting his target.
• If the polls are within the margin of error, a Republican incumbent is safe, and a Democratic incumbent is in trouble.
• And in conclusion, the Dems will NOT take control of EITHER House of Congress. Dems will lose most of the races where the polls indicated the Dems were slightly ahead. Dems will lose ALL the races where the Repub and Dem polled equal, or the Repub polled slightly higher (and within the margin of error).
• Karl Rove is right: He’s got what he calls “The Numbers”, showing Repubs holding both houses of Congress. Rove knows Republican voting patterns and tendencies. I’d bet he uses assumptions similar to those I layed out above.

Now, let’s kick back and see how I do on November 7, 2006. [UPDATE: I was wrong.]