Latest Gallup Poll Kerry up 50-44 (without Nader) 49-43 (with Nader)

[Smash255]

http://gallup.com/content/default.aspx?ci=11941&pg=1

[Lunar]

Registered voters.  Nader also got 7% here, interesting.

TIPP/IBD had some good news for Bush and now Gallup has good news for Kerry.  Both are also good news in comparison with slightly earlier polls, showing a turn for the better.

Gallup showing anyone 6 points ahead is still good news of course.

[The Vorlon]

Wow....

Gallup and TIPP usually agree almost perfectly..  (Their methodologies are damn near clones of one another in terms of data collection, the screen differently however, might be in there somewhere?)

Bush improved in the ARG, TIPP, and Zogby, but now goes down in Gallup...

I really don't know what to say about this one...

Bush job approval actually went up a couple to 49%

...this is confusing....

I will read more..

[Nation]

Kerry's campaign doesn't seem to be picking up any steam, but at least he has some numbers in his favor.

[Smash255]

Kerry's campaign doesn't seem to be picking up any steam, but at least he has some numbers in his favor.

Kerry might be getting some more steam, Bush's approval actually went up 2 points, but Kerry's lead increased.

 I haven't seen the breakouts of Dems, & Reps for this poll, but Gallup usually tends to have a 3 point Dem advantage in their polls as far as identity goes (which is pretty accurate to the country itself, 2000 exit polls shwoed a 4 point dem advantage in voter identity, and other polls suggest its in that 2-4 range) the IPP poll had Reps with a 2 point advantage in their poll.  So that could explain the differences in the polls

[The Vorlon]

Kerry's campaign doesn't seem to be picking up any steam, but at least he has some numbers in his favor.

Kerry might be getting some more steam, Bush's approval actually went up 2 points, but Kerry's lead increased.

 I haven't seen the breakouts of Dems, & Reps for this poll, but Gallup usually tends to have a 3 point Dem advantage in their polls as far as identity goes (which is pretty accurate to the country itself, 2000 exit polls shwoed a 4 point dem advantage in voter identity, and other polls suggest its in that 2-4 range) the IPP poll had Reps with a 2 point advantage in their poll.  So that could explain the differences in the polls

(Tipp had 2% more Dems than Reps BTW)

[Smash255]

Kerry's campaign doesn't seem to be picking up any steam, but at least he has some numbers in his favor.

Kerry might be getting some more steam, Bush's approval actually went up 2 points, but Kerry's lead increased.

 I haven't seen the breakouts of Dems, & Reps for this poll, but Gallup usually tends to have a 3 point Dem advantage in their polls as far as identity goes (which is pretty accurate to the country itself, 2000 exit polls shwoed a 4 point dem advantage in voter identity, and other polls suggest its in that 2-4 range) the IPP poll had Reps with a 2 point advantage in their poll.  So that could explain the differences in the polls

(Tipp had 2% more Dems than Reps BTW)

Oops my mistake.  Anyway I have another opinon on the difference in polls

In an incumbent Presidential election the undecided vote typically breaks to the challenger.  In the IPP poll their is a rather high undecided number because they don't push the undecided responses, the Gallup poll has a much smaller no opinon/ un decided since they push the undecided more.  
So the difference in the polls, could be the fact that the undecideds typically break for the challenger.

[The Vorlon]

Kerry's campaign doesn't seem to be picking up any steam, but at least he has some numbers in his favor.

Kerry might be getting some more steam, Bush's approval actually went up 2 points, but Kerry's lead increased.

 I haven't seen the breakouts of Dems, & Reps for this poll, but Gallup usually tends to have a 3 point Dem advantage in their polls as far as identity goes (which is pretty accurate to the country itself, 2000 exit polls shwoed a 4 point dem advantage in voter identity, and other polls suggest its in that 2-4 range) the IPP poll had Reps with a 2 point advantage in their poll.  So that could explain the differences in the polls

(Tipp had 2% more Dems than Reps BTW)

Oops my mistake.  Anyway I have another opinon on the difference in polls

In an incumbent Presidential election the undecided vote typically breaks to the challenger.  In the IPP poll their is a rather high undecided number because they don't push the undecided responses, the Gallup poll has a much smaller no opinon/ un decided since they push the undecided more.  
So the difference in the polls, could be the fact that the undecideds typically break for the challenger.

I more or less agree, but I would phrase it a bit differently...

This was a poll of (in rough numbers) 1000 people, 900 of whome were registered, and 600 of whom were likely.

ie - 300 or so (1/3) of the Registered sample are actually non-voters who, frankly sway in the wind and react in an opinion poll to the last story they heard on the news.  - especially when the leaners are pushed as Gallup does.

The news has all been bad, hence Kerry does well.

Within the likely voter poll, Gallup, more than the others, stresses current levels of attention being paid to the race (others more heaviliy stress past voting behaviour) Given all the news out of Iraq, needless to say a lot of Republicans have at least party tuned out, and are hence less inclined to be deemed "likely".

Factoring elements like level of attention and excitement into the likely voter screen works well very close to the election, but far out it can produce some fairly exagerated swings.

Here is the Gallup trend line from 2000, it did bounce around a ton 4 years ago also..



Hey, it's one poll of about 8 or 9, even if it is a very good one...

It is indeed odd that it disagrees with the direction of the other three polls (ARG, TIPP, and Zogby) all of which showed a Bush improvement...

We shall see...

[Smash255]

Kerry's campaign doesn't seem to be picking up any steam, but at least he has some numbers in his favor.

Kerry might be getting some more steam, Bush's approval actually went up 2 points, but Kerry's lead increased.

 I haven't seen the breakouts of Dems, & Reps for this poll, but Gallup usually tends to have a 3 point Dem advantage in their polls as far as identity goes (which is pretty accurate to the country itself, 2000 exit polls shwoed a 4 point dem advantage in voter identity, and other polls suggest its in that 2-4 range) the IPP poll had Reps with a 2 point advantage in their poll.  So that could explain the differences in the polls

(Tipp had 2% more Dems than Reps BTW)

Oops my mistake.  Anyway I have another opinon on the difference in polls

In an incumbent Presidential election the undecided vote typically breaks to the challenger.  In the IPP poll their is a rather high undecided number because they don't push the undecided responses, the Gallup poll has a much smaller no opinon/ un decided since they push the undecided more.  
So the difference in the polls, could be the fact that the undecideds typically break for the challenger.

I more or less agree, but I would phrase it a bit differently...

This was a poll of (in rough numbers) 1000 people, 900 of whome were registered, and 600 of whom were likely.

ie - 300 or so (1/3) of the Registered sample are actually non-voters who, frankly sway in the wind and react in an opinion poll to the last story they heard on the news.  - especially when the leaners are pushed as Gallup does.

The news has all been bad, hence Kerry does well.

Within the likely voter poll, Gallup, more than the others, stresses current levels of attention being paid to the race (others more heaviliy stress past voting behaviour) Given all the news out of Iraq, needless to say a lot of Republicans have at least party tuned out, and are hence less inclined to be deemed "likely".

Factoring elements like level of attention and excitement into the likely voter screen works well very close to the election, but far out it can produce some fairly exagerated swings.

Here is the Gallup trend line from 2000, it did bounce around a ton 4 years ago also..



Hey, it's one poll of about 8 or 9, even if it is a very good one...

It is indeed odd that it disagrees with the direction of the other three polls (ARG, TIPP, and Zogby) all of which showed a Bush improvement...

We shall see...

Gallup does show Kerry's lead incrasing, but the other polls you mentioned (other than Zogb showed very minimal changes.  ARG stayed at Kerry +1 with Nader, and went from Kerry +3 to Kerry +2 without Nader (rally an igsignifcant change) TIPP went from Bush +1 with Nader to Bush +2 with Nader, and Kerry +1 without Nader to Bush +1 without Nader.  Also look at the date's to the previous TIPP and Gallup polls.  Both did polls in early May at that time TIPP had Bush up Kerry 5, Gallup had Bush up 1.  Since then both polls shifted iN Kerry's favor.  Although once you throw ARG in the mix it gets more confusing because they had Kerry with about the same lead then as they do now

[The Vorlon]

ARG:

May Kerry 47/Bush 44
June Kerry 48/Bush 46

Bush gains 1%

Zogby:

Early/Mid May - Kerry 47 / Bush 42
Early June - Kerry 44 / Bush 42

Bush Gains 3% (or Kerry loses might be more accurate)

TIPP:

Mid May Kerry 43 / Bush 42
Early June Bush 45 / Kerry 44

Bush Gains 2%

Gallup

May 5-7th - Bush 48 / Kerry 47
June 4-6 - Kerry 50 / Bush 44

Bush loses 7

3 up, 1 down

The polls more or less agree on Bush's support, but the disagree on Kerry's support levels..

Bush 42-45% => Average = 43.25%
Kerry 41-49% => Average = 45.00 %

[Smash255]

ARG:

May Kerry 47/Bush 44
June Kerry 48/Bush 46

Bush gains 1%

Zogby:

Early/Mid May - Kerry 47 / Bush 42
Early June - Kerry 44 / Bush 42

Bush Gains 3% (or Kerry loses might be more accurate)

TIPP:

Mid May Kerry 43 / Bush 42
Early June Bush 45 / Kerry 44

Bush Gains 2%

Gallup

May 5-7th - Bush 48 / Kerry 47
June 4-6 - Kerry 50 / Bush 44

Bush loses 7

3 up, 1 down

The polls more or less agree on Bush's support, but the disagree on Kerry's support levels..

Bush 42-45% => Average = 43.25%
Kerry 41-49% => Average = 45.00 %

And that kind of makes sense when you look at the undecided votes tending to break toward the challenger in incumbent election on how the questions are asked.  Because the fact that the incumbent only gets a small portion of the undecided vote, their will only be a small differerence in his support when the undecideds are pushed, but their will be a decent sized difference in the support of the challenger when the undecided vote is pushed as opposed to when its not pushed.  

Tou can basically make two conclusions out of this.  1 is the strong support or those who whose minds are strongly made up on who they will vote for it's very close and too close to call, but when you take into consideration those  whose minds aren't as made up Kerry is ahead.  Now granted the possibility these votes could switch does exist, but at this moment this group is sticking with tradition and primarily going with the challenger, so having the core support of decided voters as close as it is, is not a good sign for Bush.
 So generally speaking when lets say 90% of the vote has their mind made up and 10% remain undecided the incumbent would want to be up by approx 4-5% in the 90% group that has made up their mind to make up for the fact that undecided 10% will generally go in his opponents direction.  

[millwx]

The polls more or less agree on Bush's support, but the disagree on Kerry's support levels..

Thanks for the excellent rundown Vorlon!

Incidentally, I do think there is some evidence of the "undecideds break to the challenger" theory here.  Note that excluding Mr. Wildcard (Zogby), of the other three the two with the larger undecided pool show it basically a tie (within MOE); the one with the smaller undecided pool has Kerry in the lead.  Clearly, that's not all that's going on (RV vs LV, other factors, etc), but it is probably part of it (I examined this about a month ago, when EIGHT new polls were out and saw the same thing).

I would conclude from all for of these, pushing the remaining undecideds to break similar to the rest, and pushing some of TIPP's Naders to Kerry, that if the election were today it'd be about Kerry 50%, Bush 47%... maybe closer to 51-46.

[Reds4]

The thing that looks suspicious about this poll is that it shows Bush with a lead of only 4% in red states. There is no way this is the case unless almost all of the calls are going to Ohio, Florida, W. Va, and New Hampshire. I do think Kerry could be a little ahead right now, but the numbers in the red states does not seem right at all.

[millwx]

The thing that looks suspicious about this poll is that it shows Bush with a lead of only 4% in red states. There is no way this is the case unless almost all of the calls are going to Ohio, Florida, W. Va, and New Hampshire. I do think Kerry could be a little ahead right now, but the numbers in the red states does not seem right at all.

Reds4... it's worse than you think.  Red states are ones Bush won by a significant margin (more than 5%).  So, this doesn't include Ohio, Florida or New Hampshire (among others).  They classified those close states as "Purple".  I don't question this too much, since Gallup is pretty reputable.  But, you're right, this does look fishy.  Of course, as The Vorlon points out, the subgroups may be rather small... so that could explain it too... the overall poll may be fine, but the subgroups may have a large enough MOE that they're a bit questionable.

[Reds4]

You're right, I forgot that part. However, Arizona, West Virginia, North Carolina, and maybe Virginia could be weighting the red more to Kerry still because they are closer than before it seems. If you include all the states like Texas, Oklahoma, Kansas, and so on I would think there is no way Kerry could be within 4 points in the red states.

[The Vorlon]

You're right, I forgot that part. However, Arizona, West Virginia, North Carolina, and maybe Virginia could be weighting the red more to Kerry still because they are closer than before it seems. If you include all the states like Texas, Oklahoma, Kansas, and so on I would think there is no way Kerry could be within 4 points in the red states.

Quick points:

Gallup Likely voters = 600 sample size

Sub samples (red/blue/purple) = +/-200 or so => +/- 7% (14% on the "lead")

Gallup, TIPP subsamples "might" be ok (with huge MOE) however Zogby due to his, um, unique was of doing things, inherently has totally %^%ed up internals.

The last Zogby, for example, has Bush and Kerry tied among men in terms of favorability , but Kerry up slightly among men.

[struct310]

I wouldn't count on the undecideds breaking for the challenger this time around.  It could happen and then again it couldnt.  This election is fundamentally different from other elections as it will end up being an evaluation of Bushs progress on terrorism and the economy.  The economys booming right now and that should send 50% of the undecideds over.  Another half will be concerned with terrorism and its leaning Kerry right now due to Abu Gharib.  Plus Bush has defied history several times while in office(ex 2002 elections).

[Smash255]

I wouldn't count on the undecideds breaking for the challenger this time around.  It could happen and then again it couldnt.  This election is fundamentally different from other elections as it will end up being an evaluation of Bushs progress on terrorism and the economy.  The economys booming right now and that should send 50% of the undecideds over.  Another half will be concerned with terrorism and its leaning Kerry right now due to Abu Gharib.  Plus Bush has defied history several times while in office(ex 2002 elections).

I don't think the economy booming now will turn the undecided 50/50.  Look at 1996, the economy was booming then also, and even though Clinton won  by a decent amount the undecideds in 1996 still broke in Dole's favor, and Bush has many more problematic issues now than Clinton did in 1996.  You do mention the Republicans advantage in 2002 which goes against history that is true, however much of that came from Bush's 9/11 coatails were still high at that time, those coatails have worn off.  The undecided vote will more than likley break heavily to the Challenger as usual.

[classical liberal]

Bush didn't sway the 2002 House elections, bin Laden did.