Update 11/8: Final results for the 2006 midterm election can be found at CNN's site here.
Another update just to track anything that might be changing over the last few days before the election. Polling since my last update on 11/2 indicates little change in results. Mostly thermal noise as close elections fluctuate.
I am still showing the Democrats picking up the six seats they need in the Senate to take control. In the house, the Democrats pick up one fewer seat (35) than in my previous post. At the state level, Democrats are now back to picking up seven Governors' seats.
As usual, there is more detail in the continuation and the detailed race-by-race analysis is here.
On the Senate side, the only significant change in the numbers is a widening of the New Jersey race, as incumbent Robert Menendez (D) pulls further ahead of challenger Tom Kean, Jr. (R).
In the House, the only changes I have found are one race (ID-01) where the Republican has taken the lead, and two races (CO-04 and IN-09) where the Democrat has taken the lead. In all three cases, both the previous margin and the new margin are within the poll's margin of error. It is possible (even probable) that these changes account for nothing more than random fluctuation within the margin of error.
In the Gubernatorial races, Jerry Brady (D) seems to have pulled ahead of Butch Otter (R) in the Idaho race. The previous poll showed Otter up by one. The current poll shows Brady up by 6.
The only other significant change occurred in the Alaska race, where Sara Palin's (R) lead over Tony Knowles (D) has slipped from nine points to two points. Andrew Halcro (I), however, is most likely stealing votes from Palin. It is entirely possible that some Halcro voters wake up on election day and decide they would rather have the Republican win than cast their protest vote for Halcro. I tend to doubt that Alaska will end up with a Democratic governor.
To reiterate my methodology: As you may be able to tell, I am trying to inject as little of my own opinion into this analysis as possible. That is why I am allowing close races to switch back and forth between times I publish the list. The only place where my opinion matters is that I sometimes have to pick which of two or three polls I will include in the analysis. In such cases, I tend to favor polls that are (i) newer, (ii) by national non-partisan firms versus local organizations or partisan polling houses and/or (iii) favor the incumbent. Where there are three or more polls, I also try to exclude any outlyers.
Comments