Pollster.com

Articles and Analysis

 

POLL: CNN National (8/23-24)


CNN / ORC
8/23-24/08; 909 RV, 3.5%
Mode: Live Telephone Interviews

National
Obama 47, McCain 47
(July: Obama 51, McCain 44)

Obama 44, McCain 44, Nader 4, Barr 2, McKinney 1
(July: Obama 46, McCain 42, Nader 6, Barr 3, McKinney 1)

 

Comments
djneedle83:

If you 27% of Hillary supports backing Mccain don't change your tune...we know who to blame in November!!


____________________

And what? What exactly do you propose to do about it?

That's right, Obamabully. You'll know exactly who to blame.

YOURSELF


____________________

Just so no one misses it:

Sixty-six percent of Clinton supporters -- registered Democrats who want Clinton as the nominee -- are now backing Obama. That's down from 75 percent in the end of June. Twenty-seven percent of them now say they'll support McCain, up from 16 percent in late June.

"The number of Clinton Democrats who say they would vote for McCain has gone up 11 points since June, enough to account for most, although not all, of the support McCain has gained in that time," Holland said.

____________________

player:

The democrats still have time to change their mind and choose Clinton as their nominee; Of course, Dean would scream.

____________________

Basil:

A portion of those Clinton Democrats who say they'll vote for McCain are saying so for political reasons. Some would never have voted for Hillary anyway. Others are expressing solidarity with her in hopes of somehow getting her on the ticket, and are basically playing chicken with the Obamistas.

Of the second group, only the most foolish would actually side with McCain, since Hillary's policies are far more similar to Obama's than McCain's. We'll see what Hillary, Bill, Joe and Barack can do about that.

Polls that involve Hillary are always misleading. She's leading McCain by a wider margin (IF she is at all), because McCain's not running against her and is happy to have his people promote her interests. If she were in the race it would be a different story, as any but the most cynical McCainiac would admit.

____________________

Clint Cooper:

Those who are concerned about the Hillary supporters need not worry that much. Some of those folks were never going to vote for an African American anyway. I'm sorry but that's true.

HOWEVER, Hillary is going to hit it out of the park at the convention and all the winnable Hillary voters will switch to Obama along with a lot of Independents. You will see.

____________________

Fiddling around with the reported margin of errors and assuming a simple random sample, there are 474 "registered Democrats" in the entire sample of 1,023 adults or 46.3%. Among registered Democrats, there are 171 persons who want Clinton as the nominee or 36.1%. Also, there are 784 "voters" for a turnout rate of 76.6% among all adults.

These numbers are not plausible. I could have screwed up the calculations (I'm using N = (.5/1.96*MoE)^2) which correctly calcuates N = 600 when the MoE = .04). The turnout rate is way too high for the voting-age population (it should be closer to 58% for VAP), no survey firm is estimating that many Democrats in the adult population, and I do not believe that there are still over a third of Democrats who want Clinton as the nominee.

CNN, release thy crosstabs!

____________________

faithhopelove:

It will be interesting to see the Hillary-supporter numbers after her speech at the Dem convention Tuesday night. If the two candidates are truly tied, then Obama only needs a little movement in his direction here.

That said, this poll is good news for McCain. Two other national polls have also just been released. USA Today has Obama up 3 among likely voters. A Zogby "flash" poll has Obama up 2.

____________________

faithhopelove:

Worst case scenario for Obama as he starts his convention: he is tied with McCain nationally but ahead in electoral votes. The most recent polls have him holding all of the Kerry states, and flipping AK, NV, NM, IA, OH, and VA--a total of 53 electoral votes.

____________________

John:

Completely agree with you Michael about CNN releasing their cross-tabs, but I don't think P = 0.5, that is the number of democrats voting for Obama is 50% (or at least I hope not). If lets say 75% of democrats are voting for Obama, that would lead to roughly 370 registered democrats. The number of Clinton supporters would be around 150.
Of course this would be much clearer not to mention more useful if CNN not only released their crosstabs but also their questions. How did they phrase their question to identify clinton supporters. For that matter who is winning the independent vote in their poll. Apparently only what the Clinton supporters are doing now compared to what they were doing in June matters. (BTW what were the numbers for the clinton supporters in the July poll?)

____________________

djneedle83:

Hey clinton supporters...

if you vote for Mccain in November and OBama loses Virginia/Colorado/New Mexico because of this..

We will find you...

____________________

djneedle83:

if you want


a endless iraq war....
a super right wing supreme court...
roe vs. wade overturned......
offshore drilling fairytale...
extened bush tax cuts.......

vote for Mccain...

My only beef with Obama is inceaseing the capital gains tax to 28%. That won't ever happen because congress will riot over that proposal by Obama. That is his only flaw in my book.

____________________

djneedle83:

if you want


a endless iraq war....
a super right wing supreme court...
roe vs. wade overturned......
offshore drilling fairytale...
extened bush tax cuts.......

vote for Mccain...

My only beef with Obama is inceaseing the capital gains tax to 28%. That won't ever happen because congress will riot over that proposal by Obama. That is his only flaw in my book.

____________________

player:

For all of you who think that Obama's promises are plausible. Read a little reality from George Will. Oh and by the way, George Will doesn't like McCain in the least.
http://www.realclearpolitics.com/articles/2008/08/obamas_economic_fairytale.html

____________________

JustForTheRecord:

djneedle,

Thanks for your comment. I would just add that the Obama economic plan will not increase the cap gains on households than makes more than $250,000 per year, which astonishingly on 1.5% of the country, according to census bureau.

So, accepting all that you have stated, there is no reason to vote for McCain.

____________________

Lappel1:

It occurs to me that just as there is something called strategic voting, there might be something called strategic poll response. In other words, answering a pollster a strategic responder might say that she will vote for y not because she is really going to vote for y but because she is trying to send a message to x, in order to effect a strategic change in x's policies. It is well known that people lie to pollsters when they are a little bit ashamed of what they really think. Is there any knowledge in the polling community about strategic poll response?

____________________

Ooops...I wrote the formula wrong in my post above but used the correct fomula in my spreadsheet. The numbers are right, I just transcribed the formula wrong. Serves me right for posting after driving for 6 hours.

For those interested in talking about polling methodology, here's the correct forumla: N = ((1.96*.5)/MoE)^2

Upshot, this is exactly the sort of poll to be cautious about: a flash poll conducted over the weekend. One would have hoped we would have learned something after the trouble with the New Hampshire pre-primary election polling conducted the weekend following Iowa. What may factor into these results is that every campus that I passed on my drive was jammed with students returning to school. I suspect that this CNN/ORC poll undersamples young people and if it is weighted by age, then it is probably weighting up the responses of non-college young people, who may have different preferences than their peers returning to school.

CNN/ORC, we need the crosstabs, and for good measure, please release the questionnaire, weighting scheme (if any), and likely voter model. Its difficult to have confidence in a news story without the data and methods to back it up.

____________________

zotz:

I suspect that CNN really wanted to announce poll results for the VP pick. They really played it up on their pre-convention program. Another problem is the feelings of the Clinton supporters are raw at this point. This is the political equivalent to the fog of war. We need some distance before we can see what is really going on.

____________________

thoughtful:

Good morning, Denver Day 1,

I have concluded some time ago that all the cables CNN, FOX, MSNBC have become News Entertainment Channels -I do watch Fox and CNN so I can get a hang on what the establishments wants the people to think, but that's it. MSNBC lacks critique of the left so its useless for me.

How very interesting having a tied poll on the Eve of the convention, you'd never believe it. Obama loses ground because of Clintonistas going to McCain. Only in the News Entertainment business.

____________________

Timmeh:

@thoughtful

Fox lacks critique of the right, and all they do is bash the left, so why are you watching that?

____________________

Tybo:

"djneedle83:
If you 27% of Hillary supports backing Mccain don't change your tune...we know who to blame in November!!"

Obama, right? for disenfranchising florida and michigan, running a nasty campaign against all women, for the 3 am ad, for his demeaning attitude towards clinton voters.

you'd blame the man in charge... right?

____________________

Snowspinner:

I'm trying to figure out how Obama could plausibly be dis-unifying the party, and I'm having some real trouble with it. My best guess would be momentary anger about Clinton not getting the VP nod, which has regressed Obama's support among Clinton voters to primary levels.

Will that last?

One assumes not. After all, his primary levels didn't last.

____________________

George R:

Tybo:

Obama was not "in charge" of disenfranchising Florida and Michigan. That was a decision made by all candidates in support of the DNC, including Clinton. And, of course, he's just restored full voting rights to those delegations--and will still have a majority of the delegate votes. Obama was certainly not in charge of the "3 am ad"--that was a Clinton ad and was indeed despicable.

____________________

marctx:

Thoughtless:

Notice McCain beat your 45% ceiling you ageist intolerant. How you going to show you're intolerance now with Biden on the ticket?

Republican attacks will unite democratic party? Are you as stupid as Obama? We are almost all democrats on this site.

1# Liberal - 3# Liberal 08
Change you can Xerox (both proven plagiarists)
We could have nominated Pelosi to be just a little more liberal.

Democrat for McCain-Romney 08

____________________

TheVoice_99:

Wow marctx - more BS from you I see. Too bad Obama never plagiarized - considering the guy gave him permission to use his words. Mccain on the other is a plagiarizer-


http://pundits.thehill.com/2008/08/12/mccain-plagiarizes-wikipedia-%E2%80%94-more-foreign-policy-gaffes/


____________________

thoughtful:

@marctx and @Timmeh

First of all there is absolutely no cred in this poll whatsoever.

It would be nice to see the model, the crosstabs, the weighting to see how this result is fashioned.

Whilst I am not denying that there are some Hillary supporters that would sadly rather see 4 more years of Bush McCain, than the platform that Hillary herself stood on succeed, I don't think it is as large as these numbers in this poll.If it is then it is McCain who is well cooked. Are these people really going to vote for a man who does not respect women's rights and its 2008? Apart from the continuation of the direction the country is in?

Timmeh I watch Fox so i know what distortions and mis facts Marctx has been fed from New York. So I can look forward to it.

____________________

faithhopelove:

Better news for Obama from today's McCain-favorable tracker. Obama is up 4 w/o leaners, 3 with leaners.

____________________

faithhopelove:

Better news for Obama from today's McCain-favorable tracker. Obama is up 4 w/o leaners, 3 with leaners.

____________________

faithhopelove:

(The McCain-favorable tracker is Rasmussen's.)

____________________

Patrick:

Obama had every opportunity to unite the Democratic party, create a truly exciting and historic ticket, and show respect to the 18 million people (more actual Democratic votes than he got, by the way) who passionately supported Hillary Clinton. He chose to not even consider her as his running mate. He didn't ask for a single piece of paper from Clinton during the vetting process even as he was telling people "she'd be on anybody's short list". Well, screw him. I know so many lifelong Democrats who are so angry they are either voting for McCain, Nader, or no one at all. We'd much rather have 4 mediocre McCain years followed by 8 progressive Clinton years than have such an inexperienced, dishonest, race baiting, arrogant, narcissistic flip-flopper in the White House for even one day. That's the reality he created. He'll lose the election by Ohio (maybe more) and history (and the polls) will show that he shot himself in the foot by snubbing Clinton. And Obama, who should be ahead by at minimum of 15 pts in any national poll right now, will become a national joke come November. Just watch.

____________________

atreides:

What I find interesting of late is the way the polls have flip-flopped. All the “likely voter (LV)” polls have Obama ahead by 3-4 pts while many “registered voter (RV)” have it as a tie. Since LV polls are derived from RV polls, the only way for this to happen is that they are now eliminating more McCain voters than Obama voters. Formerly they were doing the opposite under the assumption that young voters and particularly AA voters would underperform. It is apparent now that they have weighted the AA vote to overperform historical patterns and maybe hit the 75% and perhaps 80% mark. There could be a couple of reasons for this. One is that they might just do this. The other is that the Obama ground game is apparently registering bunches of new voters that the pollsters can’t track.

____________________

John:

@Michael, I think you can only use that formula for when p=0.5. That is when your estimate of event A (probability of voting for canditate A) happening is 0.5. The formula I generally use is N=(3.8416*p*(1-p))/MOE^2 for the 95% confidence level, where p is the estimated probability of event A happening.

@Atreides, are you sure?, two of the recent LV polls, ABC/Post and USA today/Gallup also released their RV numbers, both of which showed greater margins to Obama in their RV samples. I think you might be putting too much weight on this CNN poll, although it will be interesting to see if the Gallup tracker(RV) continues to show a tighter race than the rasmussen tracker(LV).

____________________

brambster:

Let me do a little debunking of the PUMA affect.

If you live in a state like OK, UT, AZ, NY, MS, CA, etc., you voted for Hillary in the primary, and you feel that the nomination was stolen from her, then it really doesn't matter at all how you vote.

In 2000 I voted for Nader as a protest vote. I live in NY. Many of my friends did the same thing. Absolutely no harm was done by voting this way. If we however lived in NH or FL and did this, it would have really been our fault for chosing a statement over a better outcome. Every one of us would have voted for Gore in 2000 if we lived in a state that was contested.

When you look at the map of states that are in play, I would guess that this covers maybe only 25% of the US population. This is where PUMA's matter, and you will definitely see a smaller cross-over effect in these states.

This effect also isn't unique to just this set of circumstances. Close to 30% of Democrats cross-over in states like Kentucky in other presidential elections. These people turned out to vote for Clinton in the primary, and while Clinton might have played well enough to contest Kentucky, not many other Democrats would. This part of the PUMA effect in states like KY has existed since Civil Rights.

So in reality, you will find a smaller than average number of Clinton supporters voting for McCain in battleground states just like Nader underperformed in battleground states in 2000.

____________________

eugene:

OBAMA SUPPORTERS DO NOT WORRY OBAMA SHOULD WIN BY 5%,MCCAIN STILL ONLY LEADS THE WHITE VOTE BY 10%,THAT MEANS HE WILL LOSE,MCCAIN HAVE TO GET THAT UP TO 19% OR HE IS GOING TO GET KILL,I DONT KNOW WHY THE MEDIA CONTINUES TO IGNORE THIS,BUT WHEN THIS ELECTION IS OVER,I BET THEY WONT IGNORE IT ANYMORE,BECAUSE THEY WILL SEE HOW BIG THE MINORITY VOTE IS,THE GOP WILL NEVER BE ABLE TO COMPETE.MCCAIN IS ONLY GETTING 25% OF THE LATINO VOTE,HE IS ONLY GETTING 30% OF THE OTHER VOTE,THAT MEANS MCCAIN IS IN A WORLD OF TROUBLE BECAUSE THE WHITE VOTE IS SHRINKING.

____________________

eugene:

OBAMA SUPPORTERS DO NOT WORRY OBAMA SHOULD WIN BY 5%,MCCAIN STILL ONLY LEADS THE WHITE VOTE BY 10%,THAT MEANS HE WILL LOSE,MCCAIN HAVE TO GET THAT UP TO 19% OR HE IS GOING TO GET KILL,I DONT KNOW WHY THE MEDIA CONTINUES TO IGNORE THIS,BUT WHEN THIS ELECTION IS OVER,I BET THEY WONT IGNORE IT ANYMORE,BECAUSE THEY WILL SEE HOW BIG THE MINORITY VOTE IS,THE GOP WILL NEVER BE ABLE TO COMPETE.MCCAIN IS ONLY GETTING 25% OF THE LATINO VOTE,HE IS ONLY GETTING 30% OF THE OTHER VOTE,THAT MEANS MCCAIN IS IN A WORLD OF TROUBLE BECAUSE THE WHITE VOTE IS SHRINKING.

____________________

Tybo:

"Let me do a little debunking of the PUMA affect.

If you live in a state like OK, UT, AZ, NY, MS, CA, etc., you voted for Hillary in the primary, and you feel that the nomination was stolen from her, then it really doesn't matter at all how you vote.

In 2000 I voted for Nader as a protest vote. I live in NY. Many of my friends did the same thing.
"

same applies to obama voters.

doesn't matter if you vote or not.
not a bit

____________________

ca-indp:

WISHFUL THINKING!

NObama is on the verge of joining other Dems LOSERS: Dukasis, Gore, Kerrey.

Obama bots are desperate and you can see their writings on this blog contain more and more BS.

Using foul language for Hillary supporter s is going to help change their views. Keep on doing it and elect McCain. That is exactly what is going to happen. Dems are Professional LOSERS.

____________________

atreides:

John-

I checked RealClear and I was looking at CNN today and the Gallup Tracking over the weekend. The Gallup numbers showed 45-45 over the weekend.
In fact the Gallup for today is out (8/22-8/24) and it shows 45-45. If the USA/Gallup LV is derived from the Gallup Tracking, then their model has to have changed. In fact Rasmussen has flipped from being slightly pro McCain to slightly pro Obama. I think they realize that they can't use the 2004 weights for AA vote for 2008. In fact Ohio and Va may not be as close because of this. Do you think that the polls overstate the older voter piece because they are so reliable?

____________________

John:

Atreides, I believe that the USA/Gallup poll is completely seperate from the Gallup tracker. I remember during the primary, that there was a large margin between the two polls which Gallup was at a loss to explain.
I don't know if rasmussen has changed their weights for the AA vote, the uptick in the tracker could be changes in the weight, a bump from the VP pick or just statistical noise. Normally we could just wait for a couple more results but with the expected convention bounce we may never know.

In terms of state polls, I agree with you, that pollsters may be undersampling AA and youth turnout, both because of historical trends which may not apply this year and due to difficulties reaching these demographics. I still think that Ohio and Virginia will be very close but the likely increased turnout in these areas probably gives Obama the edge. The pollsters may also be undersampling at the national level but very few national polls give their likely voter breakdown.

____________________

brambster:

@ John and atreides

USAToday/Gallup is a different poll with different methods from the Gallup Daily Tracking poll.

Regarding Rasmussen, Rasmussen doesn't weight by AA, or other demographic groups, they weight by Party ID. During the past couple of months they have been reducing the Democratic advantage in their Party ID weighting. So today, Rasmussen should be showing even better numbers for McCain than they did 2 months ago if everything else was equal. Rasmussen hasl also shown a skew to the Republican's advantage in 2006 compared to actual results, and in 2008 compared to other pollsters. There are several good write ups on this site about this.

The Gallup Daily Tracker also skews to McCain's advantage in comparison to most other polls, and it has a very clear weekend effect which causes Obama to lose between 2 and 4 points from mid-week to weekend in almost every week going on for months. This is clearly due to their methodology of call-backs and when to include results in their tracking. Rasmussen's tracking poll is more stable because they target a party ID mix that minimizes the call-back effect.

Regardless, there are many different polls and it's difficult to say who is right and who is wrong, but the trends/averages of these polls in combination is a better method than guessing how wrong or right a single poll might be.

____________________

John:

Brambster, I am pretty sure that rasmussen do have some weighting process for race and other demographic groups, although I am not sure of their exact methods.
http://www.rasmussenreports.com/public_content/about_us/methodology.

But I wholeheartedly agree with you about using trends/averages rather than individual polls.

____________________

Lappel1:

Four mediocre years followed by eigfht progressive years? You are playing dice with history, the kind of strategy that led Stalin's Soviet Union to pull back support for the Spanish Republicans, thinking that it would be better to suffer fascism for a while, the better to bring on the real revolution to come.

____________________



Post a comment




Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR