Kerry by 1%/6% in Pennsylvania, By 6% in Ohio?

[The Vorlon]

June 24, 2004 -- Bush, Kerry in dead heat in Pennsylvania, Quinnipiac
University poll finds; Nader holds the key to keystone state[/b]
   
Pennsylvania Polls>June 24, 2004 --
     
Bush, Kerry in dead heat in Pennsylvania, Quinnipiac University poll finds; Nader holds the key to keystone state

President George W. Bush has inched up on Democratic challenger John Kerry among Pennsylvania voters and now has 43 percent to Sen. Kerry’s 44 percent, with 7 percent for independent candidate Ralph Nader, according to a Quinnipiac University poll released today.
   
This compares to a 44 – 41 percent lead for Kerry, with 6 percent for Nader, in a May 27 poll by the independent Quinnipiac (KWIN uh-pe-ack) University. Bush's job approval also has moved up from an all-time Pennsylvania low of negative 41 – 55 percent May 27 to negative 45 - 53 percent today.

http://www.quinnipiac.edu/x11543.xml

This is the classic good news / bad news poll for both sides.

Good news for Kerry:

Now leads by 6 in the head to head ballot question( a gain of 3%)

Bush Job approval still only 45%

Good News for Bush:

Now trails by just 1% in 3 way race, up 2%
Job approval gained, went from -14 (41/55) to -8 (45/53)

Technical Notes:

Sample is a bit too republican, maybe slide everything a point or two Kerry's way...

Questions on Presidential race were waaaay down the questionaire.  The further down the questionaire a question is, the less reliable it becomes.  No way to know what, if any, consequence question order has on results, but just flagged it just in case

Unrelated quirk:

Democratic Republican Senator Arlen Spector has a higher approval rating among Democrats than he does among Republicans.



June 24, 2004
Kerry Maintains Lead Over Bush Among Likely Voters in Ohio

John Kerry continues to lead George W. Bush among likely voters in Ohio according to a survey by the American Research Group. A total of 49% of likely voters say they would vote for Kerry if the presidential election were being held today and 43% say they would vote for Bush. A total of 2% of likely voters say they would vote for Ralph Nader and 6% of likely voters are undecided. These results are unchanged from May.

In a race between just Kerry and Bush, Kerry is at 50%, Bush is at 44%, and 6% are undecided. In May, Kerry was leading Bush 50% to 43%, with 7% undecided.

http://americanresearchgroup.com/ohg/

This is actually a point better for Bush than the last ARG survey.  The last ARG showing Kerry +7 was quickly contradicted by a bunch of other polls, so I am fairly sure the last ARG was wrong.

Is this ARG also wrong is the same way the last one was, or is this one catching a new trend?

As I have previously posted, ARG is more likely to show the GOP "summer fade" than other polls...

Let's wait and see what the other polls say about Ohio...

[??????????]

You know what we need!

[agcatter]

Another Ohio Mason Dixon Poll would be nice.

[The Vorlon]

Another Ohio Mason Dixon Poll would be nice.

The "Plain Dealer" stepping up to the pump with the big $$ to do a 1500 sample size with a good firm was an unusual action of Journalistic integrity.

I don't read the plain dealer on a regular basis so I can't comment on the rest of the paper, but using M/D makes me think more, rather than less, of the folks running the show over there.

[ATFFL]

ARG has consistently had Kerry up in Ohio, and is the only service to have that.  Rasmussen had Kerry up in May, and Bush up in June.  The LA Times poll also had Kerry up, but we can ignore that.

Penn has been within a point for the last 4 polls now.  SOunds liek a toss-up to me.

[agcatter]

Yeah, seems like you mentioned once before a poll like that probably cost the paper 30 or 40K.  That clearly indicates they at least want to get it right.

[TeePee4Prez]

I can not see Nader getting 6 in PA.  

[TeePee4Prez]

I'm also thinking Kerry still has the advantage by about 4-6 when all is said and done.

[12th Doctor]

Another Ohio Mason Dixon Poll would be nice.

The "Plain Dealer" stepping up to the pump with the big $$ to do a 1500 sample size with a good firm was an unusual action of Journalistic integrity.

I don't read the plain dealer on a regular basis so I can't comment on the rest of the paper, but using M/D makes me think more, rather than less, of the folks running the show over there.

The Plain Dealer is a Liberal rag.  I read it to get a perspective on what the other side thinks.

[millwx]

I can not see Nader getting 6 in PA.  

This is a common effect.  Third party - like Nader - always polls higher (esp. this early) than they get.  Nader was pulling these numbers in 2000 in the summer as well.  He ended up around 2-3% in the actual election.

Traditional wisdom would say that these Nader-faders will go to Kerry.  However, there is a counter-argument as well... Some of these Nader-faders just won't vote.  Likewise, as most polls show Bush losing 1-3% with Nader thrown in (though, I should add, not this Quinippiac poll), some of the Nader-faders might actually go to Bush.

Nonetheless, with Kerry doing better in nearly every head-to-head poll than three-way poll, people would be fooling themselves if they didn't think more of the "Nader-faders" will go to Kerry than Bush.  So, probably, some result somewhere in between the head-to-head and three-way results is correct.  Kerry's up, in this poll, 6% head-to-head and 1% in a three way race.  I'd estimate that, if the election were today, Kerry would take PA by 3-4% (and Nader would pull about 2%, maybe 3% at most)... assuming all else in this poll is correct, of course.

[The Vorlon]

For what it is worth, my take on Nader voters is it's another word for "undecided"

What I do with them for modelling purposes is as follows:

Nader to Kerry => 50%
Nader to Bush => 20%
Nader to Nader/Don't vote => 30%

This is based upon the 2000 exit poll surveys of actual Nader voters.

By using the 2000 exit poll data I can "pretend" what I am doing has a scientific basis.

In reality, who knows for sure

[millwx]

By using the 2000 exit poll data I can "pretend" what I am doing has a scientific basis.

Vorlon... actually, I do something very similar, so I like your method  :-)

I do something similar with the actual declared undecideds as well.  No, not just because I'm trying to help Kerry out at every turn.  But, rather, in many polls, the undecideds actually go UP when Nader is added (it didn't in this poll, but it day remain constant).  That tells me that, with Nader on, some voters can't decide between Nader or Kerry or, less likely, Nader or Bush.  So, the undecided pool has better options for Kerry.  For example, if the undecided are 10% undecided between Bush-Nader, 30% undecided between Kerry-Nader, and 60% undecided between Kerry-Bush.  And if each of those pools split evenly (which the Nader pools are less likely to do - further helping Kerry), then the breakdown of undecideds works out to be... Kerry 45%, Bush 35%, Nader 20%.  And, like I said, those numbers are probably pro-Bush & Nader.  So, I use a similar ratio for the undecideds, but rather than 50/20 Kerry/Bush, I go more like 50/30.

[The Vorlon]

By using the 2000 exit poll data I can "pretend" what I am doing has a scientific basis.

Vorlon... actually, I do something very similar, so I like your method  :-)

I do something similar with the actual declared undecideds as well.  No, not just because I'm trying to help Kerry out at every turn.  But, rather, in many polls, the undecideds actually go UP when Nader is added (it didn't in this poll, but it day remain constant).  That tells me that, with Nader on, some voters can't decide between Nader or Kerry or, less likely, Nader or Bush.  So, the undecided pool has better options for Kerry.  For example, if the undecided are 10% undecided between Bush-Nader, 30% undecided between Kerry-Nader, and 60% undecided between Kerry-Bush.  And if each of those pools split evenly (which the Nader pools are less likely to do - further helping Kerry), then the breakdown of undecideds works out to be... Kerry 45%, Bush 35%, Nader 20%.  And, like I said, those numbers are probably pro-Bush & Nader.  So, I use a similar ratio for the undecideds, but rather than 50/20 Kerry/Bush, I go more like 50/30.

We are bopth taking semi-similar guesses.

We'll know November 2nd if your guess is better or worse than mine

[© tweed]

I'm also thinking Kerry still has the advantage by about 4-6 when all is said and done.

More like 1-2%

[millwx]

We'll know November 2nd if your guess is better or worse than mine

One thing for sure is that this "interpretation" is critical.  With so many close states, several flip to Kerry as compared to taking the poll numbers straight-up.  I'm hoping it DOES remain this close by Nov 2nd, so that all these theories can be verified (well, they can be even if it's not close, but it's tougher... for example, in an expected landslide, some people may just not show up... something not factored well into this interpretation).  At this stage, if nothing changes (though it probably will!), Kerry supporters have to depend on this interpretation, while Bush supporters will disregard them without "proof".

[John]

BUSH HAS A CHANCE TO WIN IN PA & HE WILL WIN IN OHIO