A new poll showing former Republican governor Charlie Crist utterly destroying former state CFO Alex Sink, a Democrat, in the 2014 Democratic primary is at best an outlier.
That's according to a cursory glance of the results from a nearly unknown polling outfit that hails from Crist's hometown, St. Petersburg, and other surveys concerning Crist that have far different results.
Bottom line, the basic methodology of the survey released yesterday from St. Pete Polls has a major oddity: It has no undecided voters in a theoretical race two years away while everyone's focused on the current presidential and (to a lesser degree) Senate races. With zero undecided votes, the poll showed Crist earning a whopping 61 percent of the vote in a Democratic primary with Sink earning just 25 percent. A handful of lesser-known candidates, including only-announced candidate Nan Rich, were in the single digits.
"There's no one undecided? Well, we've now made history: two years before an election before nearly all the candidates have really declared, everyone knows who they'll vote for," said Tom Eldon, pollster SEA Polling & Strategic Design, who has surveyed for Democrats and trial lawyers since 1996 in Florida. Along with a Republican-leaning pollster, Eldon used to conduct surveys for The Miami Herald.
Eldon was hired last month by the Democratic firm EDGE Communications to survey 600 registered Democrats to gauge their interest in the race. It found Crist and Sink basically tied, with Sink earning 31 percent and Crist 29 percent of the vote. A handful of other candidates were in the double digits. Undecided: 26 percent.
That's still great news for Crist (who's still an independent after leaving the GOP) and not-so-good news for longtime-Democrat Sink, whose 2010 loss to GOP Gov. Rick Scott rankles some Democrats to the core so much that the party is willing to seriously consider a former Republican as their nominee.
"We did this to see if Crist could compete. Resoundingly, the answer is yes," Eldon said. "Will he win? I don't know about that. There's the Jim Greer mess. And the Republicans are already running a commercial where Charlie is calling himself a 'Jeb Bush' and 'Ronald Reagan' conservative. They're putting points behind it. That makes a much bigger difference than Charlie's speech at the Democratic National Convention, which almost no one watched and that no one's talking about."
Indeed, here in Miami, the anti-Crist spot has aired enough times for two nonpolitical Miami Herald readers to note it. Neither saw Crist's convention speech. And some Democratic insiders are revolted by the prospect of Crist, while others are happy to welcome someone who has shown he knows how to win statewide.
At the least, Eldon said, the GOP ad and the Crist speech (and even his appearance with President Obama days later in Florida) probably cancel each other out. So the race is still more like Sink 31%, Crist 29%, which is well within the error margin.
There are differences in other methods between the SEA poll and the St. Pete Poll. The latter is a live-caller sample conducted by a well-established firm. The latter isn't. It was a so-called "robo-poll" where people respond to a pre-recorded voice by pushing their key pad to register responses. SEA's sample was 600 and conducted over three days; St. Pete's was reportedly 1,689 conducted on one day and of Democrats in each of the 120 House Districts. That's an interesting (and exhaustive) methodology that's as remarkable as it is commendable.
SEA polled in early August, before the convention. St. Pete polled after the convention. Also, SEA's survey was of registered Democratic voters. St. Pete claimed it was of "likely" Democratic voters. How someone can be a likely voter two years out is also remarkable.
While robo-polls are bashed by polling purists, they can be done well. A good example: Public Policy Polling, which typically surveys for Democrats. It hasn't polled the 2014 Democratic primary, but a survey last week of this year's likely voters from Public Policy Polling found that Crist's favorable-unfavorable ratings are good but not great: 44-33%.
St. Pete Poll's Crist fav-unfav: 73-16%. Gulp.
(post has been updated)