Polls Apart

[Geee]

American Spectator:

Polls Apart
By Larry Thornberry on 8.12.10 @ 6:09AM

TAMPA -- The latest Rasmussen robo-call telephone poll, released this week, shows conservative Republican former Florida House Speaker Marco Rubio leading Florida Governor Charlie Crist (ILD-Independent Leaning Democrat) for the open U.S. Senate seat by 38 to 33 with likely Democratic nominee Kendrick Meek trailing at 21 percent.

If Jeff Greene should be the Democratic nominee, the lineup is Crist 37, Rubio 36, and Greene 20. The poll has a margin of error of four percent, though Scott Rasmussen says it's a little early for polls in this competitive race to mean much.

Those keeping up with the dramatic and volatile Crist/Rubio rumble cannot help but notice that Rubio has always done better in Rasmussen polls than in any other. Sometimes considerably better. Currently other outfits -- Quinnipiac, NYTimes Newspapers, Public Policy Strategies, Florida Chamber of Commerce -- show Crist leading by from six to 11 points.

This difference has always been there. Last winter and spring while Crist was still trying to convince Florida Republicans that he was Reagan, Part II and that he should be the Republican Senate nominee (now he's Harry Truman redux), Rasmussen was the first to report that Rubio had a lead in that race. That got a few laughs from the liberal Florida media and other political "experts" who considered Crist invincible. But Rubio's lead soon began to grow to the point it was obvious even to editorial writers and political consultants. Obvious especially to Crist, who dropped out of the Republican race in April and is pursuing the Senate seat without a party affiliation. The laughing stopped.
Rasmussen was also the first to predict Joe Sestak over Arlen Specter in Pennsylvania this year. In fact, Rasmussen has a solid record of predicting winners, a record that's available on the Rasmussen website. You could look it up.

"Rasmussen and Pew were the most accurate polls in the 2008 election," said University of South Florida political science professor Susan McManus, who knows everything about Florida politics worth knowing (and probably a lot that isn't). This inconveniences those who argue that Rasmussen polls should be ignored because they overstate the strength of Republican and conservative candidates.
So why the big difference? Is a personally conservative Scott Rasmussen just painting a rosy picture of the prospects of his own kind of candidates, or does something else account for the difference? I asked Rasmussen his own self.

Scott Rasmussen didn't deny being conservative, but did laugh when I said some call him just a Republican cheerleader.
"The Republicans were complaining about us in 2006 because our numbers were showing the Democrats would win the Senate," he said
 Rasmussen said his firm's good record in picking winners and in spotting trends well before Election Day is based on solid measuring techniques, not on his outlook.

"Most of the other polling firms are generally polling registered voters rather than likely voters," Rasmussen told me. "We have a screening process that asks respondents about their voting history and their level of enthusiasm."

Rasmussen concedes that predicting who will in fact vote in an election is tricky, and more difficult than it has been in the past. He also said most firms will move to the likely voter approach after Labor Day. As to who respondents will vote for, he lists geography, race, age, gender, and party affiliation as the keys to getting an accurate sample.

"Measuring voter enthusiasm and intent is challenging, especially early in the campaign,' Rasmussen said. "But we should be a lot more confident about our numbers in October."

While denying he is a Republican cheerleader, Rasmussen pointed out some factors that should give comfort to Republican candidates in general and Rubio specifically.

"If you look at polls of generic congressional preference, Democrats do well in polls of all adults, Republicans do better in polls of registered voters, and better still among likely voters," Rasmussen said. "Groups most supportive of Democrats, young and minority voters, are less likely to be registered and less likely to go the polls."

Rasmussen said the youth turnout wasn't even that impressive in 2008, with a rock star at the top of the ticket. He said he'd be "shocked" if the youth vote turned out this year. He pointed out that in 2008 McCain won among voters who were 40 or older. Rain or shine, these folks show up to the polls, and they lean Republican.