Noting that IVR polls (polls conducted via an automatic computer voice over the telephone) in Tennessee have been consistently more favorable to Harold Ford than have polls conducted with an actual human asking the questions over the phone,
Mickey Kaus wonders if there is a new, hidden type of Democratic voter in this election:
A robopoll, Instap. notes, is supposed to minimize this Political Correctness Error because fewer people will embarrassed in front of a machine. A machine isn't going to call them "racist."
But then why the difference in the polls? Maybe a new and different kind of PC error is at work--call it Red State Solidarity Error. Voters in Tennessee don't want to admit in front of their conservative, patriotic fellow citizens that they've lost confidence in Bush and the GOPs in the middle of a war on terror and that they're going to vote for the black Democrat. They're embarrassed to tell it to a human pollster. But talking to a robot--or voting by secret ballot--is a different story. A machine isn't going to call them "weak." ...
If Red State Solidarity Error exists, it means Dems might do a bit better than the non-robo polls indicate--not just in Tennessee, but in other states where the dominant culture is proudly conservative. ...
Intrigued by this possibility
via a diary by MyDD user Tom, I went and checked out the differences between regular phone polls and IVR polls over the last month in other "red states" with competitive Senate races this year:
- Arizona: The last two IVR polls show an average lead of 49-41 for Kyl, exactly the same as his average lead in phone polls.
- Missouri: The theory holds here, with Democrat McCaskill up 46.7%--45.0% in the last three IVR polls, but down 46.3%--44.7% in the last three phone polls. However, that could also be explained by the phone polls generally being older, since McCaskill has had a bit of mo' recently.
- Montana: Democrat Jon Tester leads by an average of 51%-43% in the last two IVR polls, and by 47.5%--42.5% in the last two phone polls. That is a difference of 3%, so, the theory seems to hold in Montana. Then again, as with Missouri, the difference could be chalked up to the older phone poll being less favorable to Tester, who has been pushing outward lately.
- Ohio: The last two IVR polls in Ohio have down Democrat Sherrod Brown ahead by an average of 50.5%--41.5%. That is a pretty striking lead, especially since the last three phone polls have only shown Brown ahead by an average of 47.0%--42.7%. the difference is doubled.
- Virginia: The last two phone polls in this race have shown Republican Allen leading Jim Webb by an average of 44.5%--42.5%. However, the last five IVR polls have shown an average Allen lead of 49.2%--43.8%. Allen actually leads more in the IVR polls than he leads in the phone polls. Maybe that is because it has become so embarrassing to tell an actual person these days that you support George Allen.
So, the theory does not hold in Arizona at all. In Missouri and Montana it holds, but could be chalked up to other factors, such as the time the polls were taken. It holds in Ohio and Tennessee, but in Virginia, it is reversed.
I don't think that there is enough evidence here to conclude that there is a hidden Democratic vote in Tennessee and other red states, where conservative voters don't want to admit in public that they have at least temporarily given up on Republicans. The differences between the polls just is not large enough--it could simply be random sampling error at play. As Mystery Pollster notes:
First, let's talk about random sampling error. If we assume that all of the polls in Tennessee used the same mode of interview (they did not), that they were based on random samples of potential voters (the Internet polls were not), that they had very high rates response and coverage (none did), that they defined likely voters in exactly the same way (hardly), that they all asked the vote question in an identical way (close, but not quite) and that the preferences of voters have not changed over the course of the campaign (no again), then the results for the various polls should vary randomly like a bell curve.
Mystery Pollster goes on to note that the polls in Tennessee kind of look like a bell curve anyway. All in all, it seems very unlikely that there is a "hidden" Democratic vote out there in red states. It seems far more likely that Harold Ford and Bob Corker are in a very close race. In 2006, we would be unwise to hang our hats on unsubstantiated theories regarding a possible boost to the Democratic vote.
