Exit Polls and GOTV
Betsy's Page
Is Karl Rove just blowing smoke? The Republicans are already telling the media that they're seeing great numbers from their GOTV efforts. Early voting and absentee ballots are, they claim, close to matching their efforts from 2004.
<<< In the fiercely contested New Mexico district held by Republican Heather Wilson, the party says that the number of absentee ballots already requested by Republicans has almost reached the number requested in 2004 -- nearly 22,000 so far this year, compared with almost 24,000 in 2004. The party says it is on a pace to exceed 2004.
Meanwhile, in the bellwether Ohio district held by Republican Steve Chabot, about 60% of all early votes are coming from the roughly 40% of the electorate that the party has targeted for early voting. That's the highest rate in the country, according to an internal party memo, and good news -- "provided they vote the way we predict," the memo adds.
In two Florida districts that are in doubt -- the 13th and 16th, previously held by Reps. Katherine Harris and Mark Foley, respectively -- Republicans are ahead in both absentee balloting and early voting. In previous election cycles, Democrats have enjoyed an advantage in early voting, party operatives say.
And in the Georgia congressional district now held by Democrat Jim Marshall, "nearly twice as many of our supporters are voting as they should be, based on population," according to the internal memo. President Bush will visit the district today to stump for the Republican challenger, former congressman Mac Collins. Republicans hope to unseat Democrats in two districts in Georgia as a way of offsetting losses expected in districts across the Midwest and in the Northeast. >>>
The Democrats, of course, pooh pooh these claims and say the numbers aren't as great in other districts as the Republicans are claiming. Since early and absentee voting are a major percentage of the overall voting, the Republican claims could be an indicator of the Republican GOTV success.
<<< Experts predict this year that 19% to 25% of the electorate will vote early at the polls or by mail-in absentee ballots. That compares with about 14% in 2002, the most recent midterm election.>>>
If the Republicans are indeed successfully getting their supporters out to the early voting and absentee voting, remember that there will be a smaller percentage of Republicans actually voting on November 7. And that means that the exit polls will have a hard time reflecting the actual voting electorate. So, prepare yourself for lousy exit polls one more time.
In an effort to avoid the debacle of having their exit polls broadcast all over the internet as soon as they're delivered to the networks, they are taking special efforts to keep the numbers secret and not reveal anything until the polls close in a state.
<<< In a change from the 2002 midterm elections, the networks promise they won't project a winner in any state until after all of that state's polls are scheduled to close. In a first, each network has also agreed to send two representatives to a "quarantine room" at an undisclosed location in New York City to comb through exit-poll data. The goal: to prevent early exit-poll data -- which is often unreliable -- from leaking to the Internet, and to monitor the results in a vacuum, without access to a bank of TV screens tuned to various pundits predicting outcomes. >>>
I don't have much faith in that exit-poll quarantine room. But even if the polls don't leak out, the networks will still use that data to call elections when the polls close. And that could get very iffy in some of these tight races, because they will be missing all those early and absentee voters who didn't answer exit polls. Remember, that was one of the problems in Florida 2000 in calling the race - they missed all the Republican absentee voters. So, we could see some networks calling elections that might change once the early votes come in. But, that won't matter because there will be some who will always believe the exit polls if those polls favor their candidate.
This concerns more than a mistaken call of an election - it could also affect turnout in some of the voting districts in the western states. This is something that Karl Rove is reportedly already worried about. I would like to think that the statisticians poring over the exit poll data are aware of the hole in their data and will hold off on their calls. But I have no such confidence in their good sense.
betsyspage.blogspot.com
online.wsj.com
online.wsj.com
corner.nationalreview.com |
