« BACK  |  PRINT

RS

FRONT PAGE CONTRIBUTOR

More Quinnipiac contrarianism in Florida

By: Neil Stevens (Diary)  |  August 19th, 2010 at 01:30 PM  | 

RESIZE: AAA

Florida

Tired of Florida yet? Too bad! Just about every public pollster is hitting the state comprehensively, so like a hanging chad this race will stick around at least until next week. And after Quinnipiac yesterday went contrarian on the primaries, today the firm goes contrarian on the general.

I call this new poll contrarian because it effectively projects no difference in the general elections no matter how the major primaries go:

Charlie Crist (I) Marco Rubio (R) Democrat
Margin of Error is 3
39 32 Kendrick Meek 16
40 32 Jeff Greene 15

Yes, we do see a slight loss for the Democrats when Greene is the nominee, which is the direction I project, but the single point of movement is inconclusive. The apparent voter indifference to the candidates is even easier to see in the Governor’s race:

Alex Sink (D) Republican Bud Chiles (I)
Margin of Error is 3
31 Rick Scott 29 12
33 Bill McCollum 29 12

It’s odd to see the Republican and Independent drawing precisely identical support levels in two matchups like that, but it happens. It’s as though the Quinnipiac survey is causing those surveyed to treat the primary candidates (between both Republicans for Governor and both Democrats for Senate) as generic partisan candidates.

But time and again, history has shown that specific candidates matter in specific states for specific races, showing an ability to win or lose when a generic partisan candidate would do the opposite. So honestly I’m skeptical of this poll having any predictive value.

From Unlikely Voter.

Tags: , , , , , , , , , , , ,

COMMENTS

POSTING RULES  |  LOGIN TO COMMENT  |  REGISTER TODAY
  • rdelbov

    not give a partisan ID but they do break the vote down between the three candidates within the D/R/I catagories.

    This poll list Crist 45%D-22R-50%I with Rubio at 6%D-68%R-25%I

    So I put that into my spreadsheet at 40-40-20(roughly registered voter numbers)-but not near CNN exits in 2006 of 39R-36D-25I and I get 37% Crist and 35% Rubio. So at 40-40-20 its roughly a 2% edge for Crist

    To get 39% Crist and 32% Rubio you need to do 37%D-34%R-29%I–when you round things up you get close to a 7% Crist lead. Those are 2008 exit numbers

    So you need +3D in a registered voter poll in Florida to have Crist at +7.

    As noted it was +3R in 2006 while it was +3D in 2008 (37D-34R-29I).

    Besides it being a registered voter poll there is way too many democrats. We are seeing a huge swing ot the GOP in likely voter models and this poll shows a 2008 turnout model. 2008 for Florida was better in voter mix then 2000??

    I label this poll as trash for that reason alone.

  • Darin_H

    It could be that the GOP governor’s primary is dragging down the rest of the ticket. We’ll have to see how it shakes out after Tuesday and when the GOP electorate coalesces around the GOP primary winner and Rubio.

  • erp617

    Anybody at all believe that Crist and Chiles, son of the late “beloved” Lawton Chiles who had a lock on every office Florida has to offer for 40 years, are independent.

    Both are in the race to peel off independents, bigots and RINO’s so the Dem can win and it looks like it may work.

  • ceili_dancer

    Rdelbov is looking at the sample spread, or at least inferring it out of the results. The suppression of Republican numbers should not be a part of Quinn’s sample, it would be inserting too much into the model without going to a likely voter model.

  • rdelbov

    when I see several polls with one has Crist 39-Rubio 32 then the other has Rubio 39 Crist 32. You wonder what gives?

    Its about two factors. The partisan breakdown of the sample size and the vote breakdown for each candidate within that partisan breakdown.

    Quinnie has partisan breakdown for its sample that apparently matches Florida 2008. Why does this matter? If you polled Florida and got a sample that was 100% republican I guess Rubio would be way ahead.

    In 2008 the Florida party breakdown was 37D-34R-29I
    In 2006 the breakdown was 39R-36D-25I
    In 2004 the breakdown of voters at the polls were 41R-37D-22I

    So if you poll using these breakdowns you get varied results. Quinnie apparently is using the 2008 model-ridicious every poll shows republicans are eager to vote while democrats are depressed.

    I might add with a registered voter poll 50% of the people you poll will not vote this year.

  • tjpeco

    I just moved to Florida a week ago. Unfortunately I come to find out that I’m represented by none other than the infamous Alan Grayson. I live right down the street from a teacher’s union building that has nothing but Grayson signs out front. I need to get with the Long campaign to get some yard signs of my own.

    I’m hear a bunch of stuff on the radio waves about the two republican candidates for Gov. I honestly can’t decide who to believe. One of them might be a dirty Attorney General, the other a dirty Medicare Reimbursement scammer.

  • muffin

    So sorry you have Grayson for a rep. He’s totally certifiable. Best way to check out McCollum and Scott is to go to their websites:

    http://billmccollum.com/

    http://www.rickscottforflorida.com/

  • Dave_in_Fla

    So for now I’m taking all the polls with a grain of salt, even where I like the results.

    They aren’t just showing trends from one candidate to another, they are showing wholesale flips of support. I was scratching my head at the Scott/McCollum poll the other day, not because I didn’t think it made sense for McCollum to be recovering a bit, but because it was about a 15 point reversal from other pollsters.

    Believe me, McCollum’s ad that is running right now about supporting the AZ law isn’t good enough to reverse the damage done by the Scott ad with Bill on tape saying he doesn’t.

    Momentum changes because of events, and there have been no events to change the landscape. The ads are about the same as always, and there are no new news articles about either candidate.

    Similarly, Sink was getting killed by Scott or McCollum in the general, but she puts up one really lame ad, and suddenly gets a 15 point shift? These polls aren’t passing the smell test.

    If we get a bunch of polls from Rassmussen that confirm the trend, then I’ll buy it. Until then I’m skeptical.

    Good news is we will know how well the Quin polls are on Tuesday. If Scott wins, we know they are crap.

  • Dave_in_Fla

    Very similar levels of enthusiasm between the parties in Florida to the 2004 election.

    Oh, and those 22I? They are going to break hard to the R this time, not the typical lean toward the D.

  • rdelbov

    I might add that I also think as Bush got 15% of democrats that Rubio will also do much better then 6%–that what Katherine Harris got in 2006 among democrats. Rubio will do better then that

  • Fla Mom

    I’m not enthusiastic about either one. A friend pointed out Scott’s reply about abortion in cases of rape or incest (supports) on the Christian Coalition Florida Voter Guide. I was confused since they got the same ‘A’ rating from Right to Life PAC, but she called the Scott campaign and they confirmed. I found on ‘politifact’ that McCollum’s position on stem cells is nearly identical to Obama’s, so in terms of life issues pick your poison.

    Fla Mom