Democrats will get their stronger candidate if Kendrick Meek wins the Florida Senate primary tonight as expected- but the biggest winner coming out of the primary may be Marco Rubio. PPP finds he would begin the general election in the lead at 40%, followed by Charlie Crist at 32%, and Meek at 17%. If Jeff Greene were somehow able to pull off the upset tonight it would be much closer with Rubio at 37%, Crist at 36%, and Greene at only 13%.
PPP's last poll of the race in mid-July found Crist in the lead at 35% to 29% for Rubio and 17% for Meek. Two major developments have shifted the race in Rubio's direction though. The first is that Democrats are now going for Meek 39-38 where before they were going for Crist 44-35. As Democrats have gotten to know Meek over the course of the primary campaign they've generally decided they like him and that's cut into Crist's support for the general election.
The other big difference is that many Republican voters have moved off the fence and they've almost universally moved into the Rubio column. Where Rubio had a 54-23 lead with GOP voters in July, it's now increased to 69-20. Many Republicans were up in the air between Crist and Rubio previously but whatever they've seen over the last month has moved them more firmly into the Rubio column.
Crist's support continues to show an awkward balance that may ultimately make victory for him impossible. 57% of those planning to vote for him if Meek is the nominee think he should caucus with the Democrats in the Senate if elected while 28% think he should side with the Republicans. He's more likely to find the additional support he needs to get elected from Democrats than Republicans, but can he do that without losing the 20% of Republicans who are still with him? Whether he finds a way to thread that needle or not will probably determine his fate.
For now Rubio's in his strongest position since Crist quit the Republican race and decided to run as an independent.
Full results here
Why did you forget that you switched to likely voters and that most of all made your numbers similar to Rasmussen?
ReplyDeleteTom:
ReplyDeleteI am not sure about your analysis. If Dems have really moved to back Meek, why is he still at 17%? Seems to be Dems are instead unsure who to support. Also, Crist's support among Dems - 44 or 38 - seems to be within MoE of each other. Similarly for Meek's Dem support at 39 or 35.
Just read your report - the turnout composition seems to have a greater effect than any of the supposed movement indicated in your blog post.
Finally, if anything, Rubio's surge seems predicated on GOPers uniting around him, as you indicate, as well as a decrease in unaffiliated likely voters, as the PDF suggests.
My $0.02!
You seem to have forgoten the little thing of switching to likely voters....
ReplyDeleteI just can't see Crist winning this as he'll be squeezed from both sides by the party bases. There just aren't enough moderates that will support Crist.
ReplyDeleteThe Dems voting for Crist will want to have a definitive answer that he will in fact caucus with them (or at least support most, if not all, Dem stances on issues) and the Reps who support him will want the same.
Last month's sample was even on party id and McCain +2 on 08 vote. This month's is R+2 on party id and McCain +3 on 08 vote. LV shift accounts for little of the change in the numbers.
ReplyDeleteArggh, I want raw numbers. I hope you're doing more DailyKos polls soon, that Missouri poll was awesome- not so much the results, but the possibilities for playing around with the numbers.
ReplyDeleteHow do the Crist Republicans break down on the Scott/McCollum primary, assuming that you drew the primary voter samples from your overall sample and didn't just poll twice?
Huckabee/Rubio 2012
ReplyDeleteYou heard it hear first : )
You also forget to post about Rubio's stunning turnaround in his favs, from -15 to +3, now of course write about how the Tea Party killed GOP chances and how Rubio is unpopular, but oooops... Rubio is more popular than Crist and Meek....
ReplyDeleteDid you also poll the Governor race and if yes, when will you release it ?
ReplyDeleteAll these early polls are BS. A Quinnipiac poll (and they are one of the best of a sorry lot), four days before the CT primary had Ned Lamont ahead by four points. He only lost by 16. Only the diehards are focusing on politics before Labor Day.
ReplyDeleteCrist's numbers were temporarily inflated by his pandering to the anti-drilling fad that has mercifully passed. When Crist called a special session to pander to the anti-energy nuts, he proved that he was just another leftist nut... and that there was only one reasonable man running for the senate, Marco Rubio.
ReplyDeleteThe importance of intelligent voters:
ReplyDeletehttp://3.bp.blogspot.com/_iosqTaPsuOQ/THNDz5x60aI/AAAAAAAABwY/F3PnQgo9JAg/s1600/image001%5B1%5D.jpg
Tom,
ReplyDeleteTo what do you attribute the 12-percentage point shift in ideological composition from July (Mod +7) to August (Con +5)?
"Did you also poll the Governor race and if yes, when will you release it ?"
ReplyDeleteYes, of course. Tomorrow.
Anonymous:
ReplyDelete"You also forget to post about Rubio's stunning turnaround in his favs, from -15 to +3, now of course write about how the Tea Party killed GOP chances and how Rubio is unpopular, but oooops... Rubio is more popular than Crist and Meek...."
It really does seem like there are two kinds of Tea Party candidates: those who are like Angle and Paul and regarded by many as extreme, and those who are like Rubio and Toomey who generally aren't seen as being all that radical.
I suppose that's why Marco Rubio voted for cap-and-trade in the Florida House.
ReplyDeleteTom, the point about LV polls vs. registered voter polls is not simply the +2 point difference that you assert above. The point is that you are polling a different sample of the population.
ReplyDeleteIt has to be considered that the people who are considered 'likely voters' in midterms are people who more representative of 'base' voters - that is people who vote in midterms consistently have more party affiliation.
Your analysis may be true, it may be partially true, or it may be entirely false. In any case, you give no data that supports the analysis.
Congrats for nailing FL primaries! Gives your polls more credence! Q is really screwing up this cycle! Wrong in CT and FL.
ReplyDelete