Get yer polls, bets, numbers & pretty graphs! Elections 2008

There's some first polls that were conducted after the results in SC were known. Prepare yourself for having your feet planted firmly back on the ground: there's bad news from Oklahoma, mixed news from California, and bad news from Florida.

No bounce in Oklahoma

...when compared to a poll by the same pollster two weeks ago:

Survey USA, 1/27/2008

44% (-1) Clinton
27% (+2) Edwards
19% (n.c.) Obama

Enormous gender gap: Clinton here overwhelmingly gets the female vote, 58% to 19% and 18%, while she's in second place among men, who opt for Edwards over her by 36% to 29%.

Surprisingly, among age groups Clinton does best among young voters, of whom she gets 49%.

Obama gets 60% of the modest black vote, but only 14% of the white vote.

Obama does best among self-described liberals, a small group in Oklahoma, with 30%. Both he and Hillary do better the more liberal the voter is, while Edwards and Undecided do better the more conservative the voter is. Among the quarter of self-described conservatives in the sample, Edwards gets 33% to Hillary's 28%, Obama's 14% and 25% for Other/Undecided.

Clinton does well among the 38% of voters who cite the economy as the biggest priority, among the one in six who cite health care, and the small group who cites Social Security. Edwards does well among the small group who cite immigration. Obama does well among the one in six who cite Iraq and the small group who cites education.

When asked whether they had already made up their mind or could still change it, Hillary respondents turned out to be most firm in their choice, while Edwards' support is softest. You'd say that the smart Obama strategy would thus be to appeal to Edwards voters, but considering the political/demographic leanings of that group (white, male, older, conservative, more religious, pro-life and more likely to cite immigration or terrorism as priorities), that doesnt seem like the most promising road.

In short, if Obama's strategy is indeed to focus on the red, inland states while Hillary scoops up New York, New Jersey and perhaps California, this Oklahoma poll spells trouble (and trouble that I'm not particularly surprised about, obviously).

In fact, should the Edwards vote collapse in the wake of SC (no sign of that here yet, but it might still come), that looks like it would greatly benefit Hillary in a state like this.

Incremental change in California

Survey USA, 1/27/2008

49% (-1) Clinton
38% (+3) Obama
9% (-1) Edwards

Obama is catching up on Clinton, but is he doing so fast enough? In two weeks, the gap between them narrowed from 15% to 11%, but it's just a week until the elections.

Obama's primary problem here is that, although he is going up, Hillary is already close to snatching half the vote. This is the third poll in a row (after a Gallup poll on 1/23-26 and a CNN/LA Times poll on 1/23-27) that gives her 49%, which is considerably more than any earlier poll other than the two previous Survey USA ones had given her.

On the bright side, this Survey USA result of 38% for Obama is also better than any previous poll has polled him - so far his average this month had been 30.8%. So he is clearly going up, and if the goal is not to win the state outright but to snatch a large minority of delegates under a proportional or district-based allocation system, that's good news.

The crosstabs suggest some trouble though. Early voting is big; almost a quarter of respondents had already voted, and they'd gone to Hillary over Obama by 56% to 32%. There's simply fewer people left to be persuaded.

The gender gap is HUGE. Men actually prefer Obama over Clinton by 47% to 36%, while women go to Clinton 2:1 over Obama, by 60% to 30%.

Other patterns conform to earlier results: Obama does best among the young, Clinton among the old; Clinton does better among union members, Obama worse; Obama does well among Indies and cross-over voters, Clinton among Dems. The former only made up 13% of the sample, so this is one avenue that's still open to Obama: mobilise more independent voters, who are not allowed to vote in the Republican primaries this year.

In a race where Edwards plays a marginal role, conservative voters clearly go to Clinton over Obama (56% to 35%), suggesting again that in states where Edwards does still play a big role, like Oklahoma, he is taking away votes from her rather than him.

Numbers by race are interesting! They offer reassurance to those concerned, after SC, about Obama's ability to win white votes. In this CA poll, he actually almost ties Hillary among whites: it's 43% for Clinton, 41% for him. However, Clinton gets her large lead from two other groups: Hispanics (65% to 28% for Hill) and Asian-Americans (53% to 31% for Hill). Obama on the other hand leads among blacks, by 61% to 25%; judging on SC results there's some room for growth there still too.

What Obama could also still benefit from is the dynamics of relative turnout. This poll has 29% Hispanics, a striking 14% Asian-Americans, and 9% Blacks. If, as might be expected, turnout among Hispanics is low as it traditionally is and turnout among blacks is high the way it was in SC, he could make up a lot of ground with that alone.

Again, Hillary does well among the one in three voters citing the economy as their top concern - and I'm guessing that topic will only increase in priority. Obama does well among those citing Iraq, health care or education. Edwards does well among those citing the environment or immigration - probably reflecting his demographically split appeal to liberal activists on the one hand and conservative whites on the other.

Obama ties Hillary in the liberal Bay Area (42% each), but trails Hillary by 13-23% in the other regions.

Hillary making last-minute gains in Florida

Hillary's last minute gambit of calling for Florida and Michigan to get delegates after all is working; or perhaps it's her much-publicized last-minute fundraising events there.

Results compared to those from 4 days earlier:

Survey USA, 1/27-28/2008

52% (+5) Clinton
28% (-2) Obama
13% (+1) Edwards

Hillary's victory is all the more secure here since she gets 66% of the about a quarter of respondents who have already voted. Clintons' supporters also are most likely to say their mind is made up.

Interesting: barely any gender gap here.

Clinton gets the white vote 52% to 24%, and the Hispanic vote by a whopping 80% to 8% (sorry Soz). Obama is polling blacks 66% to 29%.

Hillary's support in this state increases the more liberal the voter is; Obama's is evenly spread among ideological orientations. Conservative Dems are again more likely to say "Other" or "Dont know"; they are the politically homeless in '08, apparently.

The economy is the top issue for a whopping 41% of respondents, and they go to Hillary more than the average, by 55% to 26%. Those citing health care go for Hillary in even larger numbers. Those citing Iraq are more likely to opt for Obama - he trails by 'only' 13 points among those.

I remember from Iowa that the bounce took a couple of days to appear. Mark Penn said "look, no bounce!" from polls taken the day after, and then whoops, the bounce (a big one!) appeared a bit after that.

Not saying that a bounce WILL appear, but I remember that was an issue with Iowa. (Someone asking Bill about that was what started his "fairytale" rant.)

Yeah, I think I remember that too.. plus there's the Kennedy event that the media went ga-ga over.. so lets wait and see.

Only prob might be that Hillary's win in Florida might dampen the momentum a bit again before it even appeared. No idea how much impact the FL vote will have. How much network media coverage is it getting, relative to the SC primaries?

Most of the coverage I've seen has centered on the delegate issue -- will they count or not?

I was watching CNN just a bit ago (I have it on in the background a lot now, same room as the computer, so I catch bits and pieces between working/ A2K'ing/ whatevering) and they said had a poll on their site about whether the votes should count. The anchors were clearly disapproving, but the poll, when I found it, was 80% for, 20% against. Since then (and after the disapproving anchors) the numbers have changed -- last I saw was 65%/ 35% I think. (Not a scientific poll of course, though it didn't seem to let me vote more than once. ;-))

http://www.cnn.com/CNN/Programs/american.morning/

But what I'm taking away from that and from comments here is that I'm very concerned that people are not "getting" the issue, and that it might go Hillary's way at the end of the day.

This is what I mean when I say the wording of polls seems like such a nebulous issue. On the one in question:

"Should the Florida Delegates Count?"

I mean, hell. There's lots of ways to take that. I'm an Obama supporter and I think they should count - but they won't b/c they broke the rules. So a big boulder of salt with polls like that.

As I said earlier if the delegate count on super Tuesday breaks out 60/40 Hillary to Obama, I'll be pleased. She can win by large margins and he can still be right in it thanks to proportional appropriation!

Cycloptichorn

Well but there's automatically a big boulder of salt... it's an online poll! It doesn't have any of the safeguards of a "real" poll.

It demonstrates the wording problem nicely though, yeah.

When the anchors talked about it they added to the barebones "yes" and "no" options -- something like, "Yes, the delegates should count even though everyone agreed that they wouldn't, or no, it's not fair if the delegates count."

Nimh,



I'm not so sure about this. We've seen record turnout everywhere across the nation in Dem primaries. I know that here in the Bay Area Obama has been running commericals and hitting the streets, I see organizers every single day out getting votes for him.

Given that Obama is more tied up in the bay area - and I wonder about LA/Hollywood and SD too - I wonder if we could see another situation in which the proportional handout of delegates gives him structural advantages. I wonder if the fact that the college educated tend to crowd in big cities, everywhere, will help him across the nation on Tues.

That's going to be a hell of a night, can you imagine?

Cycloptichorn

I dont know how that works, I'd love to know more about it.. what I remember from either Iowa or Nevada or both, I dont know which one exactly anymore, is that the allocation of delegates actually overrepresented rural and Republican-leaning counties, and underrepresented high-population, Democratic-bulwark ones. But I never quite wrapped my head around how it worked exactly, and it's probably going to be different again in CA.

Sometimes it kills me, that choosing a voting system unique to each state is about the last true expression of state's rights that we see here in America.

Some further info from Cali Progress report:



It seems that the candidate who wins the at-large vote is given a slight advantage as they will probably have slightly more delegates to begin with - and then are given more, b/c they won the vote. On the other hand, if we see a situation (like in other states) where Hillary wins the popular vote but gets less delegates, this will work to smooth that out.

It seems likely that Hillary will win but Obama will still grab a significant number of delegates....

Cycloptichorn

Last-minute prediction for the Florida primaries:

1. Romney
2. McCain
3. Huckabee
4. Giuliani

In the Dem race, eh, Hillary wins (duh)..

Zogby says I'm wrong - and he's sounding terribly persuasive:

Weather may end up being a big factor in California on election day. Southern California is reportedly mostly for Hillary and the weather is predicted to be sunny and a high of 65. Northern California is mostly for Obama and the weather here is predicted to be the 20th day of rain (it has been miserable) and high temps of 50.

We'll have to really work hard to get the vote out in NorCal.

Some unknown factors at the moment are:

According to a local news report last night about 30% of registered democrats have requested absentee mail-in ballots. Their ballots have to be in the mail by Friday to be counted. Postal date marks are ignored. They have to be in the hands of the registrar's office on the 5th to be counted. They also don't get recognized in exit-polling when the networks are making their projections.

Interesting statistic, Butterflynet, on the 30% absentee ballot requests in the Dem race in CA. I knew it was high, but hadn't heard lately how high.
Florida, in addition to mail-in absentee ballots also offers in-person voting opportunities. Again, the number of votes cast prior to today could be significant. The story I heard was somewhat clumsily reported, so I couldn't calculate the % in the Rep primary.

That could help cushion Rudi a bit since he was the only one campaigning there a month or so ago and had not yet begun to fade as other states began the selection process.
And, yes, I think Zogby's poll could accurately predict McCain's surge.

So I will go with:
1. McCain by 4 over
2. Romney, and
3. Giuliani by 3 over
4. Huckabee.

I don't see the latter two pulling out before Super Tuesday despite being drubbed in Florida (unless they cut some sort of deal). They want to go to the convention with enough delegates to get some attention.

See yall this evening right here for the results. I'll bring the crackers and that American delicacy: Cheez-Whiz.

"Of the eight organizations [polling in Florida], only three reported the number of Republican primary voters who said they had "already voted" at the time they were interviewed: 27% by SurveyUSA, 25% by PPP and 19% by the Suffolk University poll."

Pollster.com


From the same pollster.com post: "One handicap the Giuliani faced is that the period of early voting occurred over the last fifteen days. However [..], significant erosion in Giuliani's support had already occurred by January 1, so the potential gain from early voting is limited". (Fifteen days ago, Giuliani was already down to about 20% of preferences.)


Isnt Florida a winner-takes-all primary? I.e., whoever wins the Republican race tonight gets all the state's delegates?

Giuliani has all of 1 delegate right now, so if Florida is winner take all he'll still be skint of delegates hereafter, with little chance of picking up any further ones on Feb. 5 either. He might as well quit.

The winner-takes-all element really makes the Republican race more of a crap shoot than the Democratic one. According to the neat NYT map/graph Joefromchicago posted on his thread, 10 of the Republican primaries on Feb 5 will be winner-take-all! (None of the Democratic ones is.)

Huckabee is definitively in the race up through Feb. 5. He has good enough reason to hope he'll get good scores in Georgia, Alabama, Oklahoma, and of course his home state Arkansas, to wait at least until then.

The pollster.com "endgame graph" on the Republican race in Florida.

Each dot represents an individual poll result; the red and blue lines are pollster.com's sensitive and regular trendlines.

Read also the accompanying analysis

And because variety is the spice of life, here's another graphical representation of the same race:





And here's the full table of polling results. No less than 39 polls were released about this Republican contest down in Florida within 28 days - a density that comes close to the record in pre-primary polling this year, held by New Hampshire.

Latest results on top.

Touche, Nimh, on the winner-take-all factor in the Repub party. Rudi will come out of FL with nothing and with no opportunity to gain anything in the Feb 5 primaries in his own region (NY, NJ, CT, DE). He can avoid the humiliation (too strong a word?) of not being able to carry his own state.
Thanks for the gentle correction. I guess that the most he can hope for is a speech-making role at the convention sometime before 11 pm on something other than the 1st night.

A new poll from another Southern state shows much the same breakdown by constituencies. Democratic pollster Public Policy Polling (PPP) polled Tennessee on the 28th, so after Obama's win in South Carolina:

Tennessee

43% Clinton
32% Obama
16% Edwards

Among blacks, Obama gets 60% to Clinton's 20% and Edwards' 2%. But among whites he only gets 22%, which isnt much better than how he's doing in Oklahoma. Instead, 50% of white Tennesseeans opts for Clinton, and 21% for Edwards.

As in Oklahoma, there doesnt for now appear to be a youth wave to lift Obama over the top: his support is even across ages, whereas Hillary actually does best among the 18-29 year olds.

As in Oklahoma, Hillary does particularly well among those who consider the economy ("and jobs", in PPP's question) as the most important subject. Obama again does well among those who cite the war in Iraq and education, and also (puzzlingly) among those who cite immigration. Edwards, equally puzzlingly, does best among those who cite taxes.

Those last couple of things need to be taken with a grain of salt, as the crosstabs dont indicate how large each group is, and they might be small. From how the numbers relate to the candidates' overall score however it's clear that the economy is a/the top issue.

I think its a tossup, I will pick Romney to win by a point, but it could go the other way, as realjohnboy predicts, but I don't think it will be 4 points.

An observation, McCain was all over tv, as usual, along with other Democrats outnumbering Republicans, following the Bush speech, and this could be enough to sway enough voters to McCain.

Another observation, the pundits were terrible in regard to the Bush speech, grossly mischaracterizing it, not understanding it, and basically spinning it, probably with pre-prepared talking points, and they were pathetic.

How appropriate; it was a pathetic speech.

I mean, he delivered it okay, though a little boring. At least he didn't stumble and f*ck up the pronunciation of words like normal.

But he said nothing of substance, and made many false and deceptive statements - especially about things which were designed to placate you, Okie, such as the 'earmark reform' lie he told.

Cycloptichorn