Lord knows I’ve spent the last couple of weeks dogging Kendrick Meek’s campaign for what I believe is a lackluster effort to gather the requisite number of voter signatures needed to qualify for the ballot. But I criticize out of love. Tough love, but love nevertheless.
It’s for that same reason why I need to defend Meek’s position in the latest Quinnipiac Poll, which, because it showed Marco Rubio ahead of Charlie Crist for the first time, is being cited often by even casual political aficionados.
In addition to showing Rubio ahead of Crist, the poll also showed Meek trailing Crist (by 12) and Rubio (by 9). The poll also says — and here’s the real damning part if not for Meek, then for his campaign — 60% of Democrats do not know enough about him to have an opinion.
But there is a more than slight problem with the head-to-head numbers. The poll was conducted of 1,618 Florida voters with a margin of error of +/- 2.4 percentage points. The survey includes 673 Republicans with a margin of error of +/- 3.8 percentage points.
Do the math and you can ascertain that the poll is comprised of 41.59% Republican respondents. That is a dramatic oversample of registered GOP voters, by any standard.
According to the Florida Division of Elections, Democrats hold a 42%-36% advantage in terms of registered voters.
And, according to the last round of exit polls from the 2008 election, Democrats hold a 37% to 34% advantage in terms of how voters identify themselves.
From what I can tell (and I’m not positive because the poll does not say how many Democrats and Independents are in the sample, the Quinnipiac Poll oversamples Republicans by 6% (voter registration) or 11.5% (voter ID).
(And don’t start with the argument that the oversample reflects an expected surge in GOP turnout because this is a poll of registered, not likely, voters.)
So if Quinnipiac is oversampling GOP voters AND most Democratic voters don’t know who Kendrick Meek is, I think it’s safe to assume that this poll reflects a “Name vs. Generic” dynamic, meaning the heads-up questions are really a choice of Rubio or the Generic Democrat and Crist the Generic Democrat.
Based on that assumption and the fact that this poll oversamples Republicans, or just for the sake of argument, let’s say the rest of the poll is comprised of 36% Democrats and 22% Independents (a breakdown that does not favor Meek). This is a 6% swing in the GOP’s direction. So if Meek is only down 12 points to Crist and 9 points to Rubio under these circumstances, it is very safe to assume that Meek is much closer to Rubio and Crist than the Quinnipiac Poll suggests.
Oh, and by the way, if there is, in fact, an oversample of registered Republicans in this poll, doesn’t that call into question Quinnipiac’s finding that “President Barack Obama is under water in Florida”?