The margin on this race is going to be razor thin, despite kos's pessimism. Why? Look closely at the polls, and at the turnout figures for early voting. Half of Tennessee has already voted, and the Democratic counties are turning out in 15% to 20% greater numbers than the Republican counties.
More, below.
The Polls:
The polls have been all over the place lately. The DSCC poll, using a state voter file, shows Ford ahead of Corker by 46-40. The Zogby poll, using a random phone number generator, shows Corker ahead 53-43. What gives?
Zogby pulled a sample that showed a greater percentage of people claiming to be registered than they should have, by ten percentage points, and the Zogby sample over-weighted Republican voters by probably seven percentage points. The ins and outs of this poll are discussed here:
http://blog.washingtonpost.com/...
Other polls are showing the race at a dead heat, at the registered voter level, such as Rasmussen, with a spread only developing once a "likely voter" filter has been applied. The only internal polls being released are coming from Ford, and the rumor on the Corker internals is that they also show a dead heat.
None of the public polls of which I am aware are asking whether people have already voted or not. Since half of all Tennesseans who are going to vote have already done so, this would seem to be a pertinent question.
Early Voting:
Remember that there is only one contested congressional race in Tennessee--that is in Harold Ford's old district, where Harold's younger brother Jake is running as an independent against Steve Cohen, a popular and controversial state senator. This district is the most Democratic by far in Tennessee, with a 6 to 1 Democratic edge. Shelby County, which comprises parts of three congressional districts and all of Ford's old district, voted 79,000 early in 2002, and about 235,000 total. This year, Shelby has voted over 140,000 early. Shelby County has about 15% of the state's population, but has voted about 17% of the total early votes. By comparison, in 2002, in the contested Senate election, Shelby had only 13.6% of the early voting, and 14.5% of total votes in the election. It is entirely possible that Shelby County alone will give Ford an extra two percentage points.
Any analysis of Tennessee which does not take into account the changing nature of the electorate is flawed.
A detailed analysis of Tennessee early voting trends can be found here:
http://www.tnpoliticsblog.com/...
Tennessee has had over 840,000 voters vote early. This is over half of all who voted in 2002, when Tennessee had hotly contested races for governor and senator. Remember, that these early voters have already voted. They are no longer "registered voters" or "likely voters" in any pre-election poll--they are post election voters. They are nearly half of the total electorate. Democrats have an 18% turnout advantage, roughly, thus far in early voting over Republicans.
I assume that the public pollsters do not ask whether a voter has voted, but that the internal polls have done so. And I also assume that this is why Corker dumped all that money into the race, even though half of all of the votes are already in the bank. Note that Corker's ads are all negative now, while Ford's are positive-- Corker is trying to suppress Democratic turnout, not boost his own, while Ford is continuing to work to increase the already sizable Democratic lead in turnout.
When you look at this race, remember that actions speak louder than polls.