« OH-Sen: Hackett Out | Main | TX-28: League of Conservation Voters Endorses Rodriguez »

Tuesday, February 14, 2006

FL-Sen: Harris Makes Up Ground

Posted by DavidNYC

A new poll from Rasmussen (likely voters, Jan. in parens):

Nelson: 49 (54)
Harris: 40 (31)
(MoE: ±4.5%)

Now, that's a monster leap. Harris went from being down 23 points to being down just 9 in one month. Rasmussen's explanation is that the GOP has stopped trying to push Harris from the race and that big-timers (like Jeb) are backing her. But man, that's just way too much inside baseball - who even pays attention to that kind of crap (except us) this far out? I refuse to believe this kind of thing is responsible for a 14-point swing. (The one bit of bright news is that Nelson leads among independents, 64-18.)

When you have situations like this, you're always left with the same question: Which poll was out of whack? Without a third month's worth of data, we can't know. (I'm pretty sure the January poll was Rasmussen's first on FL-Sen.) If this newest poll is accurate, though, then this worries me. Nelson is still the favorite, but I was hoping we wouldn't have a serious fight on our hands here.

Posted at 02:45 PM in 2006 Elections - Senate, Florida | Technorati

Trackback Pings

TrackBack URL for this entry:


Aside from the favorable Montana Senate poll, Rasmussen released a gaggle of bad news today, including the cited comeback of Katherine Harris in Florida, Jennifer Granholm in a dead heat with Dick DeVros in Michigan, and Jim Talent opening up a five point lead against Claire McCaskill in Missouri. Combine this with Hackett's withdrawal from the Ohio Senate race last night and it's been a pretty depressing 24 hours. Let's just hope we get our depressing days out of the way now instead of November.

Posted by: Mark [TypeKey Profile Page] at February 14, 2006 03:55 PM | Permalink | Edit Comment | Delete Comment

On the other hand, I agree with you that the internals of this poll conflict with common sense and the overall result. Florida is increasingly Republican, but it's not Kansas. If the Republican candidate is losing among independents by a margin of 64-18, the only way they could still be within nine points of their opponent is if they were in a state that is far redder than Florida.

Posted by: Mark [TypeKey Profile Page] at February 14, 2006 03:58 PM | Permalink | Edit Comment | Delete Comment

Okay please help me here. Everyone keeps talking about how Republican Florida when i just don't see it... Where is this trend towards Republicans coming from? What data do we have to support it other than 2004 Election results.

Posted by: D in FL. [TypeKey Profile Page] at February 14, 2006 05:42 PM | Permalink | Edit Comment | Delete Comment

I think Rasmussen has to be "cooking the books" on these numbers in some way. Maybe the internals make sense.

Does anybody want to check and see what percentage of Democrats picked Harris in that poll?

If the partisan split is 37D/40R/23I, then Harris needs 15% of Democrats to be at 40%

I would guess Bush didn't even get 15% of Florida Democrats in 2004.

Maybe Rasmussen is robopolling the hell out of North Florida.

Posted by: RBH [TypeKey Profile Page] at February 14, 2006 05:53 PM | Permalink | Edit Comment | Delete Comment

Actually the partisan split in FL now favors Democrats. A poll conducted by Gallup last month showed Democrats with a 1% point lead. If it's any consolation George Allen is below the 50% mark here in VA meaning the race is competitive (as I have always thought).

Posted by: UVA08 [TypeKey Profile Page] at February 14, 2006 06:44 PM | Permalink | Edit Comment | Delete Comment

It's damn curious how Rasmussen materializes to give Republicans good news at crucial points in the cycle. The lone poll showing Steele leading Cardin ... the poll showing Swann leading Rendell, right before the PA GOP convention ... and this, a couple weeks after Jeb Bush endorses Harris.

Nelson's positives are still high (53-27), so I'd say: Wait for some other polls on the race. Like Quinnipiac.

Posted by: Gary Johnston [TypeKey Profile Page] at February 14, 2006 07:00 PM | Permalink | Edit Comment | Delete Comment

I don't get Rasmussen, but I don't jump on the "Rasmussen is cooking the books" train. I mean, sometimes they put out some polls that we trumpet--like Allen under 50% in VA, or McCaskill leading Talent (admittedly, not in the latest one, but the poll before that). Their results, I think, are a lot more mixed those of a clearly biased firm.

Posted by: HellofaSandwich [TypeKey Profile Page] at February 14, 2006 08:27 PM | Permalink | Edit Comment | Delete Comment

The Rassmussen numbers are not out of line with what we were tracking prior to the holidays. Harris was running in the upper 30s to a hair over 40. Nelson was in upper-50s.

In Rassmussen, all it's tracking is Nelson slipping and Harris stalled out where she was prior to the holidays.

Note that Nelson's favorables are where they were prior to the holidays. He remains well liked. So, where is his slippage? Looks like splillover from national Dems actions over past couple of months. The war and the Court, where we forward-leaned and got bitchslapped.

What has happened is Nelson has lost a bit of ground. My guess it is primarily with relatively conservative Democrats rather than indies. My guess, not seeing the crosstabs, is he's leaking on national security among nominal Dems (still registered as Dems from the old one-party dominance days but trending GOP in actual voting...a fairly big junk of registered Dems in an open primary state like Fla).

So, some nominal Dems who flirted with flipping back when Bush was at low ebb went home to the GOP on the Senate race. No biggie.

An incumbent below 50% is a bit worrisome. But when his favorables are ten points higher than his vote percentage, it tells you the problem isn't Nelson. The leakage is caused by something other than Nelson.

These Feb numbers are the first good read we've had since the first of November. From the week before Thanksgiving to the middle of January, most of the polls aren't worth squat.

Bush has rebounded. The Murtha move on the war is a drag on the party nationally, and it will be even more pronounced in southern and western states.

Hate to rain on everyone's parade, but the GOP isn't as weak of folks seem to think when babbling about taking back the House and Senate and governorships.

Nelson will still dust Harris, barring a dead girl or live boy. In the end, it comes down to the candidate. National trends are worth a couple of points either way, but the lion's share of the vote is folks choosing which of the two would be a better leader of the pack.

Harris isn't breaking the 40% floor any GOP candidate can expect even if drooling and disrobing in public. When Harris gets to the mid-40s and Nelson is still below 50...then you can worry.

But I think Rassmussen's numbers are reliable on this one.

Posted by: pinhickdrew [TypeKey Profile Page] at February 14, 2006 08:29 PM | Permalink | Edit Comment | Delete Comment

I don't think Rasmussen "cooks the books" but I think his robots poll more Republicans than other polling firms. Also his approval ratings of Bush (currently 48%) seem high. Polls that use live people like Quinnipac impress me more than Rasmussen.

Posted by: AndTun1 [TypeKey Profile Page] at February 14, 2006 08:32 PM | Permalink | Edit Comment | Delete Comment

A group that I work for in Florida just(last week of January) did a big poll for a variety of issues. We polled the US Senate head-to-head. We had the race at 25+ for Nelson.

Unfortunately, I can't disclose details so take my comments for what their worth, but, rest assured, we have never been mistaken for a pro-Dem organization. I think the Rasmussen number is way off. If anything support is solidifying behind Nelson. Harris is very weak. She'll end up with 40-45% by election day just b/c of partisan voters.

As to someone's question about the Republican nature of the Florida electorate. Bill Nelson excluded, Florida is actually a very Republican state, much more so than almost any other state in the country. Yeah, that's right in the country.

Let's see here.

US Senate: 1D & 1R (and this is the high point)
US Congress: 18R & 7D
Governor & 3 Elected Cabinet: 4R & 0D
State Senate: 26R & 14D
State House: 85R & 35D (one moderate Dem just switched to the Rep).

The fastest growing counties in Florida are all very Republican counties (exurban Tampa & NW Florida).

Florida has a Democratic registration advantage, which is mainly a holdover from conservative North Florida Democrats who vote Republican except for Bill Nelson, Allen Boyd, and their local sheriff.

Florida should become less and less of a swing state over the next 3 presidential elections.

Posted by: tallynew [TypeKey Profile Page] at February 15, 2006 10:12 AM | Permalink | Edit Comment | Delete Comment

Possible problems with the polls:
1. The demographic representation is probably different between the polls.
2. The definition of a "likely" voter is a little vague at this point.

Harris's campaigning started to come to life over the past month. She is doing better with fundraising and the Republican Party has finally recognized that she will be the nominee.

Nelson's approval numbers are pretty weak for an incumbent going into a re-election cycle.

Posted by: Tim [TypeKey Profile Page] at February 15, 2006 11:41 AM | Permalink | Edit Comment | Delete Comment

Awesome analaysis.

You have to look at the Senate numbers in conjunction with the Gov numbers. The GOP Gov numbers were better in this batch than in the last Rassumussen batch. That suggests a bad sample, or a more energized GOP base generally.

Posted by: fladem [TypeKey Profile Page] at February 16, 2006 03:59 PM | Permalink | Edit Comment | Delete Comment