November 03, 2014

Q Poll: Charlie Crist 42%, Rick Scott 41%, Wyllie (7%) draws more from Crist

@MarcACaputo

Yup, it's still tied.

Quinnipiac University this morning released a pre-Election Day poll showing Democrat Charlie Crist nursing a 1 percentage point lead over Gov. Rick Scott -- making the race basically a dead heat. Quinnipiac didn't include its poll demographics in its press release. So all you polling unskewers will have to wait.

Here's the press release:

Continue reading "Q Poll: Charlie Crist 42%, Rick Scott 41%, Wyllie (7%) draws more from Crist" »

Poll-prediction guru Nate Silver gives Charlie Crist a "slight" 60% chance of winning

@MarcACaputo

Missed this a few days ago: Nate Silver, who accurately forecast the 2012 elections (including Florida) is out with a batch of predictions and, despite the tied polling and the GOP's lead in banking pre-Election Day votes in the Sunshine State, his data show the Democrat is more likely to win the governor's race.

From Five Thirty Eight:

Charlie Crist of Florida, the former Republican turned Democrat, is just slightly more likely than not to oust Republican Gov. Rick Scott from office. A slightly safer bet for a Democratic pickup is in Kansas, where Paul Davis has a 81 percent chance of beating Republican incumbent Sam Brownback.

Silver's chart shows Crist's chances of winning at 60 percent. More here

November 02, 2014

PPP: Rick Scott and Charlie Crist tied at 44 percent, medical marijuana dying

@MarcACaputo

It has been said for months that the Florida governor's race will go down to the wire, that no candidate leads outside the margin of error in any reputable poll.

Now comes Public Policy Polling to put an exclamation on that with a poll showing Gov. Rick Scott and Charlie Crist tied at 44 percent each.

"We don't believe in reporting to decimal points but the one time we make an exception is tied final polls- Crist 44.0, Scott 43.8," PPP said on Twitter. "To give you an idea of just how this close this final Florida poll came out- 526 respondents picked Crist, 524 picked Scott."

PPP also finds medical marijuana dying at 53 percent, well below the 60 percent needed to pass a constitutional amendment.

Libertarian Adrian Wyllie "is hurting Crist," PPP reported. "His supporters say they'd pick Crist over Scott 43/18. Take Wyllie out and Crist leads Scott 47/46."

A look at the crosstabs shows that PPP has a similar finding to Quinnipiac University: a big advantage that Crist has among independents, 13 percentage points in this cast. The crosstabs:

     Overall      GOP       DEM         IND
Scott  44% 81% 18% 28%
Crist 44% 12% 76% 41%
Wyllie 6% 4% 3% 13%

As has been stated in the past, PPP is liberal-leaning but has generally produced pretty on-point polls. It is a robo-polling firm, so it primarily surveys people who have landlines and thus it could lean slightly conservative owing to its technological limits (cellphone voters tend to be more liberal, etc.). Also, PPP doesn't call off a voter list; respondents self-identify. So its polls need to be pegged to self-ID polls (such as exit polls). That's a disclaimer for all you unskewers who will say the independents are too high. They're not, judging by past exit polls.

PPP's survey was 39 percent Democcrat, 37 percent GOP and 23 percent independent. If you average the last two exit polls for the last two midterms, it would yield an R-D-I of 38-36-26 percent. Apply that to PPP's crosstabs, and the race would still be well within the margin of error: Scott 45 percent, Crist 43 percent.

For the last post unskewing the polling, click here. For more on polls in this race, click on the Polls tab here. Here's the PPP poll:  Download FLResults 

 

October 31, 2014

1,800-sample poll: Rick Scott 47 percent; Charlie Crist 44 percent. RX pot in trouble

@MarcACaputo

Gov. Rick Scott is leading Democrat Charlie Crist 47-44 percent in a new poll from Democratic leaning SEA Strategic Polling & Design exclusively obtained by The Miami Herald.

Scott’s 3 percentage-point lead is still no statistically significant because it’s within the 1,800-respondent poll’s error margin. The poll has been conducted in three waves, each of which is larger than many standalone polls (background here and here).

Still, the race is likely still a tie, though this poll indicates it's leaning Scott. Quinnipiac University's poll yesterday found it more Crist-leaning.

Meantime, the Florida medical marijuana initiative appears in trouble. Support is at 55 percent, with 39 percent opposed. It takes 60 percent approval to pass a proposed constitutional amendment such as this.

Scott’s job approval is at 52-44 percent. The poll shows that 50 percent have a favorable impression of him compared to 46 percent who have an unfavorable impression. In comparison, Crist’s fav-unfav: 44-53 percent. President Obama’s: 48-51 percent.

The poll of likely Florida voters screened from a voter list has more Republican respondents than Democrats, 43-41 percent. No-party-affiliation and third-party voters are 16 percent of the poll.

The survey’s screen reflects a relatively typical mid-term election in Florida, where Democrats typically stay home in greater numbers than Republicans. So far, in pre-Election Day voting, Republicans have stayed ahead of Democrats in casting ballots, about 135,000 more as of this morning.

Once change in this final pool compared to the prior two waves: Crist’s lead among independents has almost evaporated. It’s now just 1 percentage point (39-38 percent) over Scott.

Crist also faring more poorly among Democrats (82 percent of whom support him) than Scott is among Republicans (87 percent of whom support the Republican). Scott gets 9 percent Democratic support and Crist 8 percent Republican support.

Obviously, this isn’t good news for Democrats who must now content themselves with the hope that 1) they have a big turnout for early voting on the weekend to cut more deeply into the GOP-ballot margin 2) have a bigger Election Day turnout 3) the poll’s screen of likely voters who have voted in two of the three previous major elections hasn’t picked up a significant number of so-called “sporadic voters” who don’t get through tight likely voter screens.

October 30, 2014

Polling shows Fla governor's race could be closer than 2000

@MarcACaputo

Gov. Rick Scott is winning reelection by about 2 percentage points in a major new poll exclusively shared with The Miami Herald.

Democrat Charlie Crist is winning by 3 percentage points in Quinnipiac University’s new poll.

Which survey is right?

Both are.

The results rest within each poll’s margin of error, meaning the race is essentially a tie – regardless of the poll. Every other major survey shows that. And it looks like it will stay a squeaker through Election Day, Nov. 4.

“This race is closer than we thought George Bush vs. Al Gore was before the 2000 elections,” SEA pollster Tom Eldon said, referring to the 537-vote margin that made Bush president after 37 days of disputed results, court challenges and ballot reviews.

So Tuesday is going to be a long night?

“You’re potentially talking about a long month,” Eldon said.

More here

SEA (Dem) poll: Rick Scott 46 percent, Charlie Crist 44 percent

@MarcACaputo

Gov. Rick Scott is holding on to a 46-44 percent lead over Charlie Crist, according to a new likely voter poll exlusively shared with The Miami Herald.

Scott’s 2 percentage-point lead is well within survey’s 2.7 percentage-point margin of error – like every other recent major poll in this race – making the contest a tie. The 1,300-respondent poll was conducted by Democratic-leaning polling firm SEA Polling & Strategic Design.

A Quinnipiac University poll this morning found Crist led Scott 43-40 percent, a lead that was also within the margin of error.

The SEA poll, chartered by a coalition of businesses and exclusively shared with The Miami Herald, has been conducted in two waves over the past three days. The first results, of 800 likely voters, were reported yesterday.

While Scott’s margin has held at 2 percentage points, Florida's medical-marijuana constitutional amendment has slightly slipped by 2 points, with 57 percent supporting it and 37 percent opposing.

The amendment needs 60 percent support to pass. It still could pass if the undecideds stay home.

What makes the survey from pollster Tom Eldon stand out is that he’s one of the best in Florida, he’s a Democrat and he doesn’t sugarcoat his numbers. It’s also proof that good pollsters produce good numbers, regardless of party affiliation.

Eldon produced the poll showing Crist running strong in a bellwether seat in Pasco County.

This poll shows Scott is viewed more favorably by the electorate, relatively speaking, than President Obama or Crist.

Scott’s fav-unfav rating: 49-47 percent
Crist’s fav-unfav: 45-51
Obama’s fav-unfav: 48-50 percent.

Basically, no one is liked very much. And, as noted earlier today, all the polling and ballot numbers make this look like a squeaker of a race.

Crist used to be viewed much more favorably. But then Scott in March embarked on a mammoth $70 million TV ad campaign. Much of Scott’s ads have been devoted to trashing Crist, though the Republican has called the Democrat a “mudslinger.”

And, indeed, Crist has thrown mud. But he and his allies have less money for slinging; they’ve spent about $35 million on ads, much of savaging Scott.

Also aiding Scott somewhat is the condition of the state’s economy: 40 percent say it’s heading in the right direction; 31 percent in the wrong direction and 20 percent say it’s mixed. As for Scott’s job performance, 51 percent approve and 44 percent disapprove.

Libertarian Adrian Wyllie is not a factor, getting 4 percent of the vote.

Scott and Crist get about equal amounts of their base voters; with the Republican drawing 86 percent support from Republicans and the Democrat 83 percent from Democrats. Scott and Crist each get 9 percent support from voters of the other party.

Crist is leading Scott 38-33 percent among no-party-affiliation and third-party voters.

Crist’s lead among independents could prove crucial. Quinnipiac, which identifies party ID differently, found Crist leading by an astonishing 18 percentage points – an outlier compared to other polls. Both surveys have different methodologies.

More on polling can be found in the polling tab here.

Almost 2.4m in FL have voted, GOP lead and margin-rate slightly sliding to Democrats

@MarcACaputo

About 2.4 million Floridians have cast in-person early and absentee ballots as of Thursday morning and, as in the past week, Republicans still hold an advantage over Democrats in ballots cast but the GOP's margins are still slipping a bit, in terms of percentage.

GOP lead over Democrats: 140,123, or 5.9 percent. Yesterday, the spread was 141,363, or 6.4 percent. 

Since in-person early voting began, Republicans have alternately (but barely) increased their overall vote lead but seen their advantage in terms of rate slip. That's in part because Democrats and independents (a shorthand for no-party-affiliation and third-party voters) have begun casting ballots at higher rates.

As noted in prior posts (you can find them here), independents are looming larger and larger in the race. Just this morning, Quinnipiac University released a poll showing Charlie Crist led Gov. Rick Scott by an 18 percentage point margin among independents. That's probably an outlier result. And it needs to be pointed out that Quinnipiac uses self-identification polls in which respondents tell the pollster what their party is; so some of these independents are Democrats and Republicans.

However, another poll released yesterday by SEA Polling and Strategic Design showed Crist winning independents 37-33 percent. And that poll was conducted off a voter list, so those margins can theoretically be compared to the independents who have voted so far. 

Here's what happens if you apply those independent results to the pre-Election Day ballots: Scott's potential 140,000 lead gets cut by by about 12,000 to almost 128,000, an 8.7 percent reduction.

Apply the entire SEA poll's partisan crosstabs to the early ballots, and Scott theoretically leads Crist 47-42 percent (a 133,000 margin).

A word of caution: the above calculations are just a math exercise to give one glimpse into how the race is playing out. We still don't know how people actually voted. This is also based on one poll. All polls have error margins. And all major polls recently find the race basically tied. Unlike Quinnipiac, SEA finds Scott doing slightly better with his own base (Republicans) than Crist is with his (Democrats) and it finds Scott doing slightly better than Crist among crossover voters from his opponent's party. Also, the poll has 6 percent undecided.

Here are the early and absentee ballot numbers as of this morning:

PARTY         EV          %
REP             330,497 41.3%
DEM             333,711 41.7%
IND             136,340 17.0%
TOTAL             800,548  
     
     
PARTY         AB's           %
REP             714,315 46.0%
DEM             569,327 36.6%
IND             269,937 17.4%
TOTAL          1,553,579  
     
     
PARTY      EVAB           %
REP          1,044,812 44.4%
DEM             903,038 38.4%
IND             406,277 17.3%
TOTAL          2,354,127  

NOTE POST HAS BEEN UPDATED (Some prior numbers were wrong at the margins).

Q Poll: Charlie Crist, fueled by independents, leads Rick Scott 43-40 percent overall

@MarcACaputo

Charlie Crist has slightly nudged ahead of Gov. Rick Scott in Quinnipiac University's latest poll that shows the Democrat picks up 43 percent support from likely voters to the incumbent's 40 percent.

That apparent 3 percentage point lead for Crist is essentially a tie because it's within the poll's margin of error, however the poll of 800 likely voters indicates Crist has some momentum on his side.

Last week, Quinnipiac surveyed the governor's race and found it dead even between Crist and Scott, with each getting 42 percent. So the contested has shifted a net 3 percentage points in Crist's favor or a net 5 points compared to Quinnipiac's September survey that showed Scott with a negligible lead.

Boosting Crist: the strong support of self-identified independents, who favor the Democrat over Scott by a spread of 47-29 percent.

“Independent voters are often the difference in swing states like Florida, but the size of former Gov. Charlie Crist’s lead among them is truly remarkable,” Peter A. Brown, assistant director of the Quinnipiac University Poll, said in a written statement.

Continue reading "Q Poll: Charlie Crist, fueled by independents, leads Rick Scott 43-40 percent overall" »

October 29, 2014

SEA (Dem) poll: Rick Scott 45 percent; Charlie Crist 43 percent

@MarcACaputo

Florida’s race for governor is as tight as ever, with Gov. Rick Scott getting 45 percent support and Democrat Charlie Crist 43 percent from likely voters, according to a new poll from Democratic-leaning polling firm SEA Polling & Strategic Design. 

The Republican’s 2 percentage-point lead is well within the 800 voter survey’s 3.5 percentage-point margin of error – like every other recent major poll in this race.

Like other polls, this one shows that Florida's medical-marijuana constitutional amendment is within a point of passing or failing. It shows Amendment 2 polling at 59-32 percent. The amendment needs 60 percent support to pass. Another way to look at it: opponents need 40 percent to oppose it. If the undecideds stay home, the amendment likely passes.

What makes the survey from pollster Tom Eldon stand out is that he’s one of the best in Florida, he’s a Democrat and he doesn’t sugarcoat his numbers. It’s also proof that good pollsters produce good numbers, regardless of party affiliation.

Eldon produced the poll showing Crist running strong in a bellwether seat in Pasco County.

Contrary to the belief among some in Tallahassee and other political circles, Scott has secured his base. He’s getting 84 percent support from Republicans; while Crist draws 80 percent from Democrats.

Crist is not disproportionately getting crossover voters from the opposite party. In fact, on that count, they’re basically tied. Crist gets 11 percent support from Republicans, Scott gets 10 percent from Democrats.

Crist is leading slightly among independents: 37-33 percent. That could prove crucial.

For comparison, the averages of the partisan crosstabs of 10 polls released earlier this month showed Republicans backing Scott over Crist 72-13 percent; Democrats backing Crist over Scott 74-13 percent and independents favoring Crist over Scott 34-39 percent.

This is all inside-the-error-margin stuff.

No matter how you slice these polls, or this poll from SEA, this race is essentially tied.

And, yes, it’s time to say the smartest and dumbest thing in politics: It’s all about turnout. Republicans are turning out more of their voters in pre-Election Day ballots, but Democrats have more voters.

The poll reflects a Republican turnout advantage of 2 percentage points, showing an electorate of 43 percent GOP; 41 percent Democrat and 16 percent independent.

Change those numbers and the topline numbers change.

For more polls in this race, here's our polling tab.

October 28, 2014

Internet poll from NY Times/CBS: Rick Scott 46%, Charlie Crist 45%

@MarcACaputo

Gov. Rick Scott leads Democrat Charlie Crist 46-45 percent in the latest internet poll from YouGov, conducted for the New York Times/CBS.

So this poll, like others, essentially shows a tie because Scott's 1-point lead is within the error margin.

YouGov's polling technique is controversial and experimental (some thoughts at the bottom of this post). In looking at the survey's crosstabs, it has some strange results: Scott winning independents by 12 percentage points. No recent survey (or perhaps any other survey in this race) has had such a margin. The only recent one that had Scott up was a poll from the Republican Party of Florida. Most polls have Crist winning independents by 3 to 5 points.

On the other hand, the poll appears to have over-surveyed Democrats, thereby giving Crist back some of his margins lost among independents in the survey.

The internals of the poll aside, what's interesting here is that Crist has somewhat narrowed the gap in the survey, which has consistently shown Scott doing better than most other polls.

Paradoxically, though it's an outlier, the survey toplines seem right.