There seems to be a real divergence in polling.  On one end, Quinnipiac, Fox News, and CNN/Time has the Governor’s race at a 5-7 lead for Kasich.  On the other end, you have SurveyUSA, the Columbus Dispatch’s mail-in poll, and today’s Quinnipiac poll that claims that Kasich has a 12-17 point lead.

Earlier this week, Stu Rothenberg, a highly respected political forecaster dismissed the results of recent SurveyUSA polls and the Dispatch poll, saying:

That survey, which showed Republican John Kasich with a 12-point lead over incumbent Gov. Ted Strickland (D), had an inexplicable sample that was 51.9 percent male and 87.3 percent white. In the 2006 midterms, exit polling showed the Ohio electorate to be 49 percent male and 79 percent white. That’s a problem.

Even more significant, while voters aged 65 and older constituted 34.4 percent of the Dispatch’s sample, they accounted for only 19 percent of ’06 midterm voters in Ohio. Not surprisingly given the overabundance of older voters in the Dispatch poll, younger voters are dramatically underrepresented in the recent survey.

Kasich would have been leading Strickland even if the sample looked better, but the margin would have been much closer, and almost certainly much closer to reality.

Or take the bizarre August SurveyUSA poll in Virginia’s 5th district, which showed Democratic Rep. Tom Perriello trailing his Republican challenger, Robert Hurt, by an astounding 26 points, 61 percent to 35 percent.

Democrats criticized the poll’s methodology, but I didn’t even need that to know the numbers were a joke. Both Democratic and Republican pollsters show Hurt ahead narrowly (and he is favored to defeat Perriello in November), but nothing like 26 points.

What’s interesting about the polls is that on partisan breakdown, there’s not much difference.

In today’s Quinnipiac poll of what they believe is likely voters, the partisan breakdown was this.  Strickland attracted 84% of the Democratic vote and 32% of the Independent vote.  Kasich got 91% of the GOP vote and 55% of the Independent vote.  In other words, in a poll where Kasich shows his best lead, each candidate has roughly their party base solidly behind them—Kasich more so—with Kasich sitting on a margin slightly larger than twenty points with Independents.

What does a poll like Rasmussen and CNN/Time—which are more favorable to the Strickland campaign—say about the candidates’ partisan performances?  Well, yesterday’s Rasmussen said:

“Kasich, an investment banker and former congressman, now earns 88% of the Republican vote in the state. Seventy-seven percent (77%) of Ohio Democrats support Strickland. Among voters not affiliated with either major party, the GOP candidate leads by more than 20 points.”

In other words, Rasmussen shows just about the same picture.  In fact, it actually shows Strickland performing worse with his base than Quinny.  What about the CNN/Time poll that also showed a seven-point race?  In that poll, Strickland gets 84% of the Democratic vote and 38% of the Independent vote.  Kasich gets 89% of the Republican vote, and 54% of the Independent vote.  See a pattern?

So, wait…. if Quinnipiac, Time/CNN, and Rasmussen all show essentially the same numbers for the candidates with voters based on those voters partisan views, how does Quinnipiac find a seventeen point race when the other two shows a seven-point race?  Well, because those numbers are what they produce in their likely voter model.  The only way Quinnipiac and the Columbus Dispatch gets the results they do is that their likely voter model is based on a more heavily GOP electorate than what Fox News and Rasmussen is predicting for Ohio.  Read that sentence again and again and again.

The problem with the Dispatch, SurveyUSA, and Quinnipiac polls are that they are predicting that the 2010 electorate in Ohio will be mostly Republican—in a State in which the Democrats still have a 1 million voter registration over the Republicans.

Beyond that, we already have hard data, actually physical evidence, that the fundamental assumption of Quinnipiac, SurveyUSA, and Dispatch mail-in polls are fundamentally flawed.


Franklin County is considered a swing county, ordinarily.  However, it is also consider the heart of John Kasich’s geographic base.  And in that county, just as many Democrats have said they don’t want to wait to vote on Election Day as Republicans.  That is the very thing that Quinnipiac and the Columbus Dispatch’s statistical likely voter model “predicts” wouldn’t happen.  Except it already has.

There are a million more of us than there are of them, and we’re ready to vote.  Don’t let a poll based on an already proven flawed likely voter model get you down.

Tagged with: