Pollster.com

Articles and Analysis

 

SurveyUSA: GA, NJ, NY (10/11-12)

Topics: PHome

SurveyUSA
10/11-12/08
Mode: IVR

Georgia
547 likely and actual voters, MOE +/- 4.3%
McCain 51, Obama 43
(9/30: McCain 52, Obama 41)

Senate: Chambliss (R) 46, Martin (D) 43, Buckley (L) 6
(9/30: Chambliss 46, Martin 44, Buckley 5)

New Jersey
551 likely voters, MOE +/- 4.2%
Obama 55, McCain 40
(9/29: Obama 52, McCain 42)

New York
547 likely voters, MOE +/- 4.1%
Obama 64, McCain 31
(9/25: Obama 57, McCain 38)

 

Comments
IndependentThinker:

I like the NY one (Obama +33), I am happy my home state is made up with smart people who can see the light

____________________

Viperlord:
____________________

DocnTN:

@ IndependentThinker

Wish I could say the same of my present home state of Tennessee. The McCain posters litter the entire landscape down here..

____________________

Flashlight:

Come on, Georgia!

We need a blue stripe from Maine to Florida!

Go Georgia go!

____________________

Nhoj:

all these seem about right haven't looked at the crosstabs yet though.

____________________

falcon79:

fyi:
40% of early voters in Georgia are African-American. Obama currently has a double digit lead among votes cast so far in Georgia.
If the trend continues, then Georgia may very well turn blue, in which case it could b argued that these polls coming out of Georgia are downwardly biased because many of these pollsters do not poll first time voters.

____________________

I don't think we'll see Georgia flip this year, but seeing it move to light pink is sure gratifying. Who knows, maybe McCain will have to spend time defending the heart of The South instead of trying to pick off Pennsylvania.

____________________

IndependentThinker:

@DocnTN

Hopefully that will happen someday
LOL!!

____________________

sulthernao:

Blacks make up 26% of the vote in GA? I don't think so. They will make up at least 28% of the vote (they make up 29% of the registered voters). Don't be surprised if they make up 30% of the voters.

____________________

Save US:

NY... you guys are just a bunch of dumb people... as opposed to Alabama... :)

____________________

drational:

@Falcon79
"Obama currently has a double digit lead among votes cast so far in Georgia."

You are likely close to correct, but since the ballots will not be counted until after election day, you have no idea how the ballots will turn out.

____________________

Disco Stu:

@falcon79:

Now that more voters have voted early (18%), Obama is only up 52-46 (among those already voted).

Still...having a 6 point lead with ~18% of votes in is something. Here's the link to the crosstabs:

http://www.surveyusa.com/client/PollReport.aspx?g=6a397798-d00a-4fac-ae44-cd99e38e4e6b

This poll also has GA even (41-41) in party affiliation...anyone know if that's accurate?

____________________

MileHigh:

It doesn't matter what McCain does unless he wins Florida. Check out 538. I love what I'm seeing.

____________________

Trosen:

*yawn*.. come on... Need MO, VA, NC, MT, NV, FL, and OH polls please.. now.

____________________

sulthernao:

@Falcon79

Do you have a link? Because that would invalidate the GA poll (SUSA has Obama up by only 6 among those who have already voted).

____________________

miatch:

This New Yorker loves his city!

____________________

ericsp28:

PLEASE stop posting polling information about early votes already cast. You can effect the outcome of the race by doing so. And any media outlet that releases polling data on actual votes cast before the polls close is acting highly unethically. You are only making the problem worse by posting them in other forums.

____________________

drational:

@ Disco Stu
"Now that more voters have voted early (18%)"
"Obama is up 52-46 (among those already voted)"
What in the hell are you talking about?

____________________

NW Patrick:

This New York # concerns me for McCain. He'd better campaign there!:)

____________________

falcon79:

@ sulthernao:
Unfortunately, I dont have a link, because I heard it on Countdown with Keith Olberman last week (early last week, maybe?)
However, it seems like Disco Stu's information is more current - according to Stu, Obama has a 6 percent lead with 18% of the vote in... doesn't that sound like its possible these SUSA numbers are too much in favor of McCain?

____________________

falcon79:

@ drational:
Check the SUSA internals.
Early and absentee voting began in Georgia on 09/22/08. Among the 18% of Georgia voters who tell SurveyUSA they have already voted, Obama leads by 6 points; among those who have not yet voted, McCain leads by 11.

____________________

carl29:

Anybody wants to make extra bucks in FL?

"McCain Hiring Paid Canvassers in Florida"

McCain is now offering to pay people to do door-to-door canvassing for the campaign.

An e-mail went out over the weekend from the Republican Party in Hillsborough County, which encompasses the vote-rich Tampa area, inviting supporters to go door-to-door for the campaign for $12 per hour. "Work as many shifts per week as you want," it said. "No experience necessary, but a strong desire to make a difference and a strong work ethic are important. So is reliable transportation."

:-) Go Obama/Biden!!!!

____________________

Disco Stu:

@ericsp28:

How is posting info on polls of people who already voted unethical? I guess they should post opinion polls either because those might sway voters too!?!?!

If you have a problem with it, take it up with SurveyUSA.

For those of you who were looking for a link to the SUSA GA poll, here it is again:
http://www.surveyusa.com/client/PollReport.aspx?g=6a397798-d00a-4fac-ae44-cd99e38e4e6b

____________________

drational:

@falcon79, DiscoStu

18% of the votes are not in, and these are not official results.
This is poll data from a small sample (18% of 547= 98 voters)= meaningless.

____________________

zoot:

IIRC, there were enormous numbers of unregistered AAs in GA at the beginning of this cycle. 538 has played around with data available as of 9-30 and come up with scenarios in which Obama wins, based on a combination of increased AA registration and increased AA participation. While the assumptions represent a bit of a stretch, they're worth taking a look at. It would be astounding if McCain is in any kind of real trouble there, but if his numbers are even approximately accurate, it may mean significant problems for Chambliss.

http://www.fivethirtyeight.com/2008/10/in-georgia-small-improvements-in-black.html

____________________

laguna_b:

RCP has Obama up to 50% on weighted polls average!

____________________

sulthernao:

@falcon79'

I think the problem with SUSA's polling in GA is that they are for some reason undercounting the number of black voters. They have it only at 26% (2004 turnout) when registration has jumped to 29%.

____________________

Scott in PacNW:

I like how close that Senate race is. Nothing would please me more than to see that P-O-S Chambliss tossed out.

____________________

ericsp28:

I can assure you, I have taken it up with SUSA.

When you claim that candidate A is ahead in the early voting that only encourages opponents of candidate A to vote who might not have otherwise voted. And if you are a supporter of candidate A that has not yet voted and you see that your candidate is ahead in early voting then you may decide that it is not worth the trouble to vote since it looks like your candidate is going to win.

There is a reason why the MSM won't release exit polling data until after polls close.

____________________

carl29:

sulthernao,

If early voting is any indication of the energy among AA's in Georgia, things are going to be closer than polls show them to be because roughly 40% of early voters have being AA's.

____________________

jeepdad:

Early and absentee voting began in Georgia on 09/22/08. Among the 18% of Georgia voters who tell SurveyUSA they have already voted, Obama leads by 6 points; among those who have not yet voted, McCain leads by 11.

I'm thinking Obama leads by much more than 6% among those who have voted early. He's "only" registering 89% of the AA vote in this poll. He will likely get more like 95%. This thing could end up being a 3% race.

____________________

falcon79:

@drational:
Assuming that the sample of 547 voters is a truly random sample, the sample statistics will be highly reflective of the actual population parameters.
The law of large numbers states that as sample size increases, sample statistics and population parameters converge.
Given a random sample, even a sample of 30 individuals will provide a robust estimate of the population parameter values.

The issue you raise is the size of the sample. This will translate into the estimate being somewhat inefficient, which is why there is a margin of error term.
We do not know what the margin of error for the statistic "number of people who voted" is, but extrapolating from the main MOE, it is likely that anywhere between 14% and 22% of the population has already voted in Georgia.

Hence, your criticism that "the small sample size makes this poll rubbish" is an invalid criticism, because random samples are valid to the extent that they are truly random, and not to the extent of their respective sizes.
Were this not the case, statistics would not in fact exist as a science.

How do I know all of this?
Well, for starters, I teach Statistics at a large state university...

____________________

IndependentThinker:

OMG North Dakota is now a toss up on pollster's map, does that mean McTitanic will send Palinocchio there?
It's getting funny!

____________________

Stonecreek:

RE: Polling of early voters.

How is it less ethical to ask people how they intend to vote than to ask them how they actually voted? Is a voter's opinion somehow less representative if they early vote? Wouldn't the polling results be in danger of being skewed if already-voted persons were excluded? Doesn't including votes already cast by definintion improve the Likely Voter screen's accuracy?

How can you say that an apparent advantage in early voting, when reported, tends to discourage voters who favor the candidate who is leading? Most political scientists would argue, if anything, the opposite.

I am truly chagrined that someone would raise the polling of early voters as an ethical issue. Just where did you obtain your Doctor of Ethics degree, Mr. Goody Two-Shoes? Please go wring your hands and pontificate somewhere else. Besides, do you think that anyone besides we politi-nerds and polling-curmudgeons will ever know or care what percentage of the Georgia polling results are actual early voters and how they are voting; and how many of us will be influenced in our own vote? Zilch. So get over your ethical eruption.

____________________

rami:

Well it's obvious Obama wins in NY and NJ and Mccain wins in GA... But, the figures are interesting.
Obama is +33 in NY
Mccain is +8 in GA

In 2004 (check link),
Kerry was +17 in NY
Bush was +17 in GA

So obama leads in NY is twice Kerry's lead, and Mccain's lead in GA is half Bush's lead...

But it's still extremely unlikely Obama wins GA or any deep south state (I was thinking the DNC would send the clintons campaigning in Arkansas).

____________________

carl29:

rami,

The other day Hillary was campaigning in Arkansas. Chuck Todd just mentioned that he thinks someone should poll AR.

____________________

tjampel:

It might be a good time for McCain to consider re-assigning Palin to permanent duty in N. Dakota. This will keep her FARGO(ne) from the National limelight, where she's become a serious drag on his attempts to sway independents, moderate undecided Repubs, and soft conservative Dem Obama support.

Shes definitely got the Fargo accent down perfectly (at least the Coen Bros. version of it). 3 straight weeks of Palin in N.Dakota should definitely bring state back from the brink.

____________________

Trosen:

Last AR poll was back on 9/22 and had MCCain +11. In light of recent events and poll trends, I'd also like to see some new AR polls.

____________________

CTPonix4BHObama:

@ flashlight

"Come on, Georgia!
We need a blue stripe from Maine to Florida!
Go Georgia go!"

..so im not the only one that plays tic tac toe, or bingo of sorts making continuous lines of blue states across the country.

Also, if people are going to assume that many poll numbers are wrong because of AA turnout then isnt also somewhat safe to assume that will also affect congressional races because of downticket voting?

____________________

WhereisMitt:

I agree that Obama will not GA, but surely he will outperform previous Dem candidates on election day.

Full disclosure: I'm a lifelong conservative and have always voted Republican. However, I've been disappointed with the McCain campaign since the beginning (I wanted Romney) and the Palin pick for VP left me at a loss for what to do. I might just write in Romney. I think we Republicans need eight years in the wilderness to find ourselves. If Palin is the future, then we're all in trouble.

Anyway, I know quite a few people who were among those who voted early for Obama in GA and as a Virginia native I know people who have never been motivated to vote at all (both conservative) who are voting for Obama this year. Just anecdotal evidence, but it is nonetheless a indication of the political climate out there.

____________________

IndependentThinker:

@tjampel

You're right about re-assigning Palin permanently in North Dakota
Also, I think she should stay there till Nov 4th and then take a non-stop flight to Alaska on Nov 5th and never comes back

Obama/Biden 08

____________________

burrito:

@tjampel:

LOL. I agree. She could also improve her foreign credentials, as you can see Canada (the higher populated part) from ND!!

____________________

NW Patrick:

9 min. to Rasmussen. I repeat. 9 min!

____________________

Atomique:

Woohoo! We are going to make New York the bluest damn state in the country on November 4! I♥NY!

Come on. You think you can beat us, Rhode Island? Bring it on, Maryland. We're gonna be bluer than Illinois, bluer than California, bluer than friggin' Massachusetts. OK, the District of Columbia is going to have us all beat, but New York is going to be next on the list!

____________________

[nemesis]:

Actually, you'd see the higher populated part of Canada from Michigan or Washington/Western Montana. Please send her to Montana. Michigan is far too close for comfort.

____________________

NW Patrick:

My RACIST uncle is voting Obama. That was ENOUGH SAID for me. He's 94, x military, potty mouth, jerk. But I'll take his vote.

____________________

Gary Kilbride:

Please keep pushing this Georgia theme, on 538 and elsewhere. I want lower Intrade prices for the red candidates in Georgia. In wagering you always get bargains by bucking the tide, not favoring it. In Obama's case it's like the primaries, too much spillover optimism in the unlikely states. I remember all the rationale that he could pull out Pennsylvania and Ohio. Right now Georgia is trading in the high 70s, which is far too low. But if you want to give me free short term interest of 20% and up, be my guest.

Meanwhile, this may be the only recent cycle in which New Jersey did not tease the Republicans with early phony low margins.

____________________

NW Patrick:

Isn't anyone a Ras prem memeber? Spill it!

____________________

ericsp28:

@Stonecreek

Polling people who have actually voted vs people who intend to vote is two completely different things. Pollster all have different methods of sampling the likely voter population which is why the actual numbers in the polls are usually all over the place. For example, looking at today's national polls, you can believe Obama is up by double digits, or only up by 4 depending on what pollster you believe. As an Obama supporter, I want to believe its by double digits, but I'm not going to take it for granted that the pollsters that are showing a smaller lead are wrong so I'm going to vote anyway. If I see a poll of early voters in my state that shows Obama has a huge lead, and if it was a hardship for me to get to a polling place to cast a vote, then I might decide that it isn't worth the effort.
That doesn't mean that you exclude people who have already voted from your sample though. You just don't ask them the question of whether or not they have actually voted, and if they volunteer that information there is no reason to report it.

If its not a big deal, then can you explain to me why all of the major media outlets are putting very strict safeguards to ensure that no exit polling data is released before polls close in any given state? Why would they go to so much trouble if they didn't think it was a big deal?

And by the way, I'm a business analyst by profession, so I know a thing or two about statistics, not that my credentials should have any bearing on whether or not I am capable of holding a position on the matter.

____________________

NW Patrick:

Per Rasmussen moments ago.

Obama up 49-47 in OH. (Trailed by 1 previously.)

Obama up 50-47 in VA.

Obama up 51-46 in FL.

Obama up 50-47 in MO.

Obama and McCain TIED in NC 48-48.

____________________

[nemesis]:

Haha, saw a quote today by Palin saying it was a sprint to the finish (or something like that, paraphrasing.)
I think the only sprint we can liken this to would be one in which one of the sprinters wasn't allowed to use their legs.

____________________

fed:

Ok, battleground polls are out today, Rasmussen daily tracking will go up again tomorrow, two days before the next battleground polls are out, Obamas lead shrinks again in the tracking polls. Look at the trends in the daily tracking polls and then the days that the OH and FL polls come out and you will see what I mean

____________________

muckinello:

I love it! RCO updated their map and put FL as leaning Obama ... now up to 304 EV in this most conservative site!

____________________

falcon79:

@ericsp:
you make some good points, and your capability on holding your position was never in doubt.
however, i can also see the other guy's perspective: seeing we are all polling nerds here, i doubt its a big deal if we just discuss among ourselves - i doubt too many people pay attention to out blog posts here lol

____________________

sotonightthatimightsee:


Blacks make up 26% of the vote in GA? I don't think so. They will make up at least 28% of the vote (they make up 29% of the registered voters). Don't be surprised if they make up 30% of the voters.


**cough, cough**

Excuse me, I believe your statement is racially tinged and you owe the McCain supporters an apology, no?


How did that feel? ;-)

____________________



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