Diageo/Hotline Tracking Poll: The Early Line

Obama-Biden 46%
McCain-Palin 44%
Undecided 9%
-- The race has tightened in the last day. After trailing by 5-7 pts. for the last 10 days, McCain is now just 2 pts. behind Obama.
--One potential reason: Obama's one-time lead on the question of who'd best handle the economy has evaporated. Today, Obama and McCain are tied at 42%. Independent voters favor McCain on the economy by an 8 pt. margin (42-34%).
--Even so, among the 62% of voters who picked the economy as their #1 issue, Obama leads McCain by 9 pts. - 49-40%.
Today's Diageo/Hotline tracking poll, conducted 10/4-6 by FD; surveyed 908 LVs; margin of error +/- 3.3% Party ID Breakdown: 40%D, 38%R, 18%I.
(AMY WALTER)





Can somebody familiar with the poll comment on the party ID breakdown? Why did it go from a +5 Dem advantage to +2? Seems to me that this change - which doesn't reflect any of the other party ID numbers I've seen out there - would be the primary cause of the shift from Obama +6 to Obama +2.
Who are you guys polling? I've never been polled? I'm a consistent voter? I have young friends across the country who've never been polled. I honestly do not think this poll is anywhere near correct!
You guys are not polling the youth vote!
Who are you guys polling? I've never been polled? I'm a consistent voter? I have young friends across the country who've never been polled. I honestly do not think this poll is anywhere near correct!
You guys are not polling the youth vote!
North Carolina is definitely going to Obama!!! I promise that!
Who are you guys polling? I've never been polled? I'm a consistent voter? I have young friends across the country who've never been polled. I honestly do not think this poll is anywhere near correct!
You guys are not polling the youth vote!
North Carolina is definitely going to Obama!!! I promise that!
Who are you guys polling? I've never been polled? I'm a consistent voter? I have young friends across the country who've never been polled. I honestly do not think this poll is anywhere near correct!
You guys are not polling the youth vote!
North Carolina is definitely going to Obama!!! I promise that!
Who are you guys polling? I've never been polled? I'm a consistent voter? I have young friends across the country who've never been polled. I honestly do not think this poll is anywhere near correct!
You guys are not polling the youth vote!
North Carolina is definitely going to Obama!!! I promise that!
to OGLiberl: Basic lesson is always that weird fluctuations in sampling happen too easily to put a lot of credence in one poll. You're right that the party ID split is suspect. Three other trackers out so far today - Research 2000, Rasmussen and Battleground - and they show no movement. We'll see what Gallup says this afternoon. Only when the average of the five moves, and the movement persists for a few days, do I really believe it. This Hotline shift may be illusory - we'll see if it bounces back tomorrow.
You guys will have to excuse all the Obama maniacs that go berzerk if a poll doesn't show Obama winning in a landslide. The race is going to tighten. By next week i predict a tied race again.
Why is this poll so out of step? Independents prefer McCain on the economy?!? Everything else I have read says the exact opposite.
Umm, let me second the earlier commment about party ID above.
I see your track poll released on Monday had ID at 41% D 36% R, vs. 40% D 38% R today. Can you explain why the change in sample?
Also, given the size of your sample the independent/uncommitted appears to be less than 200, so that is far from statistically significant.
I agree with OGLiberl that the party ID differential completely skews the results of this recent poll. Rasmussen is consistently polling with a 5-7 % differential between Democrats and Republicans and given the predicted higher turnout for Democrats this election cycle as evidenced by the higher turnout in the 2006 congressional elections and the 2008 presidential primary, the 5-7 % difference may be an underestimate. So my question is why did this poll weight the party ID to near parity between the two parties if not to mislead the public with the false perception that John McCain is starting to narrow the gap?
The only thing that makes me suspicious of this poll is the ID breakdown, no way that the difference between the dems and the Reps is only 2%, try more like 4-6% in favor of the dems and you'll probably be closer to the real value, that's the only reason for this swing towards McCain
the party id difference was 4-5% in 2004, but has grown significantly... its more like 7%-9% difference. so this poll is just an outlier.
This is called MSM bias. They want to make the race appear close. How do you do that, play with the party ID. Shame on you Hotline...do you really think we are that stupid. Oh well obiviously you do.
Why have your party ID numbers shifted so radically? Even Rasmussen has Dems maintaining a +6 advantage. Untilo you can explain the shift, your poll will lack credibility.
Why have your party ID numbers shifted so radically? Even Rasmussen has Dems maintaining a +6 advantage. Until you can explain the shift, your poll will lack credibility.
It is really funny how people get paid to do poll's like this. Do you guys get paid to make it tighter or something? Any sane poller knows the Party ID is not 40%D, 38%R, 18%I, it makes no sense what so ever and yet here we are.
Well, those worried about the economy in this poll favor Obama by 9%. So clearly this poll seems to have a relatively lower % of folks not worried about the economy than do most polls. that might reflect the apparently narrower margin than most polls have in partisan weight.
I echo other comments here asking for an explanation of the change in party ID. Hotline has been assuming a fairly conservative 41D 36r split. What happened to cause the revision to 40D 38R?
Calling these results evidence of a tightening race is rather strange, since your results changed entirely because of the change in party ID weights.
40D 38R, huh?
Interesting party ID.
The truest test of the accuracy of this poll is the complete panic of the McCain campaign. There is no chance that they would be going wildly negative if they were close. In addition, there is no correlation to basically every state poll. In other words, an outlier. Dems should start to worry when Palin becomes pleasant and they stop unleashing third rate sheriffs to do their dirty work.
The tightening shown in this poll doesn't seem to be reflected in other credible national polls, most of which continue to show an Obama lead of 6-10%.
I don't discount this entirely, but unless the trend shows up in other polls, we've got to consider this one an outlier, probably due to the arbitrary re-assignment of party preference weights in the polling.
That said, I still would expect that this election is going to be close - the country remains ideologically divided and the kitchen sink campaign of McCain/Palin will take a toll....but in this economic environment, that kind of fear-mongering isn't going to beat the desire of the public for real leadership and partisan change.
I have worked in research and polling my entire life. These people are correct re: the change in sampling. It may be justified, but you need to offer such an explanation or else you appear to be deliberately skewing the results or appear to be incompetent.
This poll does NOT weight by party ID, I believe, which is fine, but makes it susceptible to unusual sampling situations, like this, which undercut the actual party ID by a significant margin.
All polls are meaningful (note that I don't consider anything with a self-selected pool of respondents to be a "poll") they just need to be put into context, and the context of this one is seen in the party ID.
At least we have the data to make the interpretation. Given no significant movement in the R2k or Rasmussen polls today, this is likely a sampling anomaly and will snap back in a few days, plus or minus any underlying change.
Like the other blogger.....I have never being polled either......and not NC is not going for Obama either TRUST ME!!!! I live in a military town and believe me is 5 to 1 mccain here.....
@ vinny from NC
I live in NC as well with family spread about NC, I see Obama support overwhelming here. Party registration for NC is heavily favored for Dems in NC, and most recent polls show Obama leading.
Vinny from NC, do not underestimate the black vote. Look how many votes were casted during the primaries for both Hillary and Obama. If those folks show up to vote NC will be Obama! Do not underestimate the youth or black/latino vote this time around, we will be supporting in numbers. Not making this a racial thing but this is the first time we've ever had someone to put our issues as important as the others. Do not underestimate us! We do not get polled!
PPP Poll: Obama Expands Lead in North Carolina
A new Public Policy Polling survey in North Carolina finds Sen. Barack Obama now leading Sen. John McCain, 50% to 44%.
For the first time, "Obama is earning over 80% of the vote from self identified Democrats, and that's fueling a four point increase in his lead in the state compared to last week. He now has an 82% to 15% lead with voters in his own party. His share of the Democratic vote had been anywhere between 69% and 76% in PPP's previous five surveys of the state."
Interesting: Obama's numbers rise every time he makes a visit to the state. McCain hasn't visited in five months.
Vinny from NC, do not underestimate the black vote. Look how many votes were casted during the primaries for both Hillary and Obama. If those folks show up to vote NC will be Obama! Do not underestimate the youth or black/latino vote this time around, we will be supporting in numbers. Not making this a racial thing but this is the first time we've ever had someone to put our issues as important as the others. Do not underestimate us! We do not get polled!
PPP Poll: Obama Expands Lead in North Carolina
A new Public Policy Polling survey in North Carolina finds Sen. Barack Obama now leading Sen. John McCain, 50% to 44%.
For the first time, "Obama is earning over 80% of the vote from self identified Democrats, and that's fueling a four point increase in his lead in the state compared to last week. He now has an 82% to 15% lead with voters in his own party. His share of the Democratic vote had been anywhere between 69% and 76% in PPP's previous five surveys of the state."
Interesting: Obama's numbers rise every time he makes a visit to the state. McCain hasn't visited in five months.
Vinny from NC, do not underestimate the black vote. Look how many votes were casted during the primaries for both Hillary and Obama. If those folks show up to vote NC will be Obama! Do not underestimate the youth or black/latino vote this time around, we will be supporting in numbers. Not making this a racial thing but this is the first time we've ever had someone to put our issues as important as the others. Do not underestimate us! We do not get polled!
PPP Poll: Obama Expands Lead in North Carolina
A new Public Policy Polling survey in North Carolina finds Sen. Barack Obama now leading Sen. John McCain, 50% to 44%.
For the first time, "Obama is earning over 80% of the vote from self identified Democrats, and that's fueling a four point increase in his lead in the state compared to last week. He now has an 82% to 15% lead with voters in his own party. His share of the Democratic vote had been anywhere between 69% and 76% in PPP's previous five surveys of the state."
Interesting: Obama's numbers rise every time he makes a visit to the state. McCain hasn't visited in five months.
While Diageo may not weight the "responses" on the voter identification, they CAN filter out particular party and candidate preferences by adapting who they consider "likely voters". Unless they tell us who they consider "likely"...what algorithm they use to decide this, and who they are removing from their polling sample, I would tend not to trust them. Are they giving much higher weight to 65+ voters, to those that are resident for years in their homes, downgrading new registrants or those that "didn't vote in 2004" [automatically extracting everyone ages 18-22], etc.
BTW New Democratic registrations (none of whom are expected to vote, of course) are up dramatically in Pennsylvania (see todays news)...as they are nationwide. Polsters should ignore these numbers at their own peril.
While Diageo may not weight the "responses" on the voter identification, they CAN filter out particular party and candidate preferences by adapting who they consider "likely voters". Unless they tell us who they consider "likely"...what algorithm they use to decide this, and who they are removing from their polling sample, I would tend not to trust them. Are they giving much higher weight to 65+ voters, to those that are resident for years in their homes, downgrading new registrants or those that "didn't vote in 2004" [automatically extracting everyone ages 18-22], etc.
BTW New Democratic registrations (none of whom are expected to vote, of course) are up dramatically in Pennsylvania (see todays news)...as they are nationwide. Polsters should ignore these numbers at their own peril.
The Obama nuts are really in an uproar! Calm down!
It seems obvious that like John Kerry in 2004, Obama has peaked too soon, and is on his way down to defeat.
Try again 2012, Libs. LOL.
McCain Palin 08
Nobama....
Given the sample sizes involved, there is nothing mysterious about the change in Party ID. Hotline does not weight by ID so it will vary from poll to poll. The sample over this time period has more Republicans than would be expected compared to the average of samples Party ID of all reandom samples. This variability is an inherent issue with polling for elections.
The "likely voter" algorithm is something that almost no pollsters disclose - it's proprietary data. Most pollsters use some combination of past voting behavior as reported by the respondent, respondent's reported interest in the campaign and respondent's own estimate of likelihood of voting, but the exact blend is not reported. I have never heard of a public pollster (I used to be one - on a minor scale) cooking the formula to make things come out a certain way.
Now we know that today's poll is an unweighted outlier. Gallup just came out for today and Obama is up by 9 pts. (51%-42%), which mirrors Rasmussen today at +8% (52%-44%). So weighting party ID does matter, and in a poll that apparently does not weight by party ID, the results will be misleading. The pollster should be more honest and state up front if in fact the poll was not weighted by party ID.
I just looked at the Gallup Poll and I also think the race will tighten. Obama's camp and the pundits say the numbers are high. There are so many new factors that polls cannot be accurate anyway.
I'm a Obama supporter, but agree with the McCain supporter that polls will begin to show a closer race. Anything can happen between now and Nov. 4.
This is not a credible poll result, because of the weighting distortion. I recommend that DIAGEO should hire a well qualified Statistician to handle their polling for them.
Glad to see the race tightening.
This poll is in line with CBS and Zogby.
Polls are worthless this election.
I would not be surprised to see either candidate win with numbers far out of line of the polls at this point, too many media agencies in the tank for one candidate or the other, and these polls are just there to try and influence your opinion.
Should be banned for 60 days prior to an election just like they are in Europe to be honest
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I am pretty sure no one knows what is going to happen next let alone how to solve the present predicament. No one now believes in anything. I believe we are entering into another realm of life and anyone who has the extraordinary oratory skills and a little magic will win the votes not only in the US but globally.
I would like to know who they are polling? I am in my 40's and have voted for years and I have never BEEN POLLED!!! I think it is a joke. Almost everyone I ask says they are NOT voting for OBAMA but yet the polls say differently. This is sick and a bunch of lies!!
Mccain looked feeble last night. How is his health? Are we ready for Palin to run our country? I think not.