I’ve been in court pretty much all day today, but it looks like the media has lost its mind on polling with today’s release of the Quinnipiac poll showing Kasich at a 10-point lead. The Dispatch reported it as if Kasich’s lead is growing even though any person with any knowledge of polling would mock any such assertion when Quinnipiac showed virtual no movement anywhere.
But, heck, even polling that shows no change in the race isn’t good news for Strickland who still needs to close the gap.
But let’s keep everything in perspective.
Let’s first remember that Quinnipiac’s likely voter model has been the most generous to Kasich ever since they started using it. (In the first likely voter poll, it showed Kasich’s biggest lead ever, a seventeen-point lead that even Jon Keeling admits was flawed and cannot be believed. The only other polls that have come close to Quinny’s ridiculous numbers was the constantly mocked, yet still done anyways, Dispatch “mail” poll and SurveyUSA, which Stu Rothenberg has indicated could not be believed.
Although Quinny has come down from that number, there’s no evidence that they’ve changed their likely voter model at all. I don’t expect a soon to be released SurveyUSA poll to be any more reasonable than the last one in September that mirrored the ridiculous results that have only been exceeded by the Dispatch’s mail in poll and that 17-point lead in Quinnipiac, unless they’ve demonstrated that they’ve substantially changed their likely voter model. For that reason, the Strickland campaign has said that “Quinnipiac polls are irresponsible, inaccurate, and completely removed from the reality of the Ohio governor’s race.” And they are absolutely correct.
In short, a month later, and the same observations in this post still apply. The only reason these polls are showing Kasich with a lead is because they’re predicting more Republicans than Democrats and Independents will vote, something early voting strongly suggests won’t be the case (and since we’re half way through early vote, we’ve got a statistically reliable sampling.)
Remember, the only reason the Ohio Poll showed Kasich moving from a four-point lead to an eight-point is only because the Ohio Poll shifted its likely voter model to a more GOP electorate in which nearly half the electorate was Republicans. Had the Ohio Poll used the same model it used in September, the Ohio Poll would have shown the race actually narrowing to a two-point race, which is pretty much exactly what the latest Rasmussen poll has shown.
There is no evidence, none, that the smallest population of the registered voter population in Ohio (registered Republicans) will be the biggest population of the general election population in November to such an extent that they will be a majority of the general election vote (which is the only way a poll like Quinnipiac can show a lead greater than the Ohio Poll, which requires an unprecedented 49% GOP electorate.)
In 2008, when Ohio was considered a “tossup” State in the Presidential election, exit polling suggested only a 31% electorate. The idea that we’ve gone from a +8 Democratic electorate to a > –7 one (a fifteen-point swing in two years) is unbelievable in a State where Democrats have nearly a million voter registration advantage and is reported leading overall in early voting.
I’ve had more than one source that Kasich’s internal polling showed him only ahead by two points, even after Kasich’s pollster on Twitter insisted to us that it was five-points. Kasich is campaigning heavily in counties today that went for Blackwell. The National Review has said that Kasich has a problem in his base. His campaign is overwhelmingly negative. In other words, Kasich’s campaign and his supporters like the RGA don’t act like they believe they have this race won, so neither should you.
Keep working the phones, talking to your friends and neighbors, and we can prove them wrong on November 2nd.