There's a recommended
diary up now suggesting that the elusive cell-phone-only voter has been surveyed by Zogby and will break overwhelmingly for Kerry. This is cited as evidence that there is a large subset of voters uncounted in traditional polls and on whom we can count to put Kerry well over the top.
I will say nothing of the rather unusual nature of the poll - whether a practically self-selected sample in this case is at all representative of the population at large - nor will I argue whether or not the cell-phone only demographic is as large as some people who frequent this blog suggest. Furthermore, the rather dubious result that less than 3% of 18-29 year olds in this survey fell they will not likely vote is one I'm not going to address.
Rather, the big hubbub seems centered on the 55-40% lead Sen. Kerry enjoys over Mr. Bush among the 18-29 year olds counted in this survey. In many people's minds, this seems to validate the whole theory that an army of cell-phone only users will prove the traditional polls wrong.
Well, Zogby himself says:
"The results of this text-message poll mirror what we're seeing in our more conventional polls. Among 18-29 year-olds, Kerry leads the President by 14 points--55% to 41% in our current daily tracking poll--virtually identical to these results. Our text-message poll seems to have been validated by this experiment. All in all, I think we've broken some new ground in polling."
Great. Good for him. He's essentially demonstrated that cell-phone-only users, who may have been polled in his text message poll, will vote exactly along the same lines as those who are reached via a landline in a traditional poll. It is true that the 18-29 y.o. demographic will vote overwhelmingly for Kerry, but traditional polls already adjust for such demographic information like age. So Zogby's poll essentially says, if the traditional polls weight their age distribution correctly, then it doesn't matter whether there are a bunch of cell-phone only users uncounted in traditional polls, because they will vote the same way as the landline users.
The real argument is whether the traditional polls have the breakdown correct, or whether the 18-29 year olds will actually vote this time around at a rate greater than they have done historically. That is a separate discussion entirely.