« New AP--Ipsos Poll: Kerry, Bush Tied | Main | Strategy Notes:
John Belisarius

Democratic Unity is More Important than Yesterday’s Debates »

WSJ/Zogby Poll: Kerry Ahead or Tied in All Battleground States

On the eve of the 2nd presidential debate, Senator John Kerry is ahead or in statistical tie with President Bush in all 16 "battleground states," according to a new poll by Zogby Interactive, conducted 9/30 to 10/5 for the Wall St. Journal.

Kerry LV leads (%): AR 0.2; FL 0.4; IA 6.6; MI 9.7; MN 8.3; NV 1.0; NH 6.6; NM 11.4; OH 0.3; OR 10.1; PA 5.4; WA 9.9; and WI 2.5.

Bush LV leads (%): MO 2.2, TN 0.9 and WV 6.1.

According to the analysis of the Zogby Poll published in the Wall St. Journal, "Mr. Kerry holds leads outside the margin of error in 6 states...None of Mr. Bush's leads are outside the margin."


You got to be kidding me - Then again this is a partisan Dem site, so I guess it's understandable. Read by bytes: This is an internet poll based exclusively on Zogby subscribers. It should only be used as propaganda in a DNC blaster Email, rather than a microcosm of our electorate.

Insofar as Zogby has become more of a partisan Dem, as opposed to being an independent poll taker, Reps are less likely to subscribe to his site - And thus less likely to participate in hiw "battleground" poll.

Zogby doesn't publish internals for this poll but I bet his party ID mix is something like: Dems 50%, Reps 25%, Ind 25%. With a party mix like this, how can this poll be credible??

These polls are nice to see. But an 11.4% lead in New Mexico? When most polls show NM to be very close? Seems doubtful that Kerry could be so far ahead.

Some other encouraging news...SUSA shows Kerry up by 10% in Michigan, the Becker Institute shows Kerry up by 6% in New Hampshire. And Gallup has Colorado dead even, which suggests that we might just be able to pick that one off.

It's close in Iowa, with SUSA showing Kerry up by 1%.

The bad news is that Gallup has Bush up by 3% in New Mexico and Wisconsin.

If the only changes from 2000 are that Kerry flips Ohio and New Hampshire to his side, but Bush gets New Mexico and Wisconsin, the race ends deadlocked, unless Colorado passes its proportional bill or one of Maine's districts goes to Bush.

As someone who desperately wants kerry to win, I still have to question the accuracy of interactive polling. What do you all think? It iseems quite possible to me that people could be seeking out to participate in these polls to influence the news coverage.

Weak employment report this morning, and last month's numbers were revised down. Bush is getting slapped from all sides. Pity.

Hey, I think I'm in there. I got a Zogby Interactive email last week and got polled. And I'm sure if they were trolling for LV's, I was certainly included.

Then again, I live in Texas which isn't a battleground state (yet!), but I really think that getting polled by Zogby went more towards getting my vote counted than actually showing up on election day (though I plan to do that too).


This is excellent news. If tonight's debate goes well, and I believe it will, I think Kerry will win by about 10 points.

So now there's photos showing that Bush may well have worn a radio receiver to the first debate, in violation of the rules. Incredible.

I'm open to the argument that interactive polling will do better than telephone polling. Given declining participation rates in telephone surveys, I think there are some real questions there.

But I'll simultaneously be skeptical of the interactives till I've seen more of a track record. I'm inclined to use interactives for trends (thus I'm encouraged by the movement toward Kerry from the last Zogby poll, while being unsure whether the overall numbers are on one side or the other of truth). And when interactive state polls are confirmed by phone polls, I'm encouraged that the answer is pretty close.

When there's a conflict, I scratch my head and fall back on the "this thing is close. Keep working!" mantra.

Mara, Regarding:

"Weak employment report this morning, and last month's numbers were revised down. Bush is getting slapped from all sides. Pity."

"Weak" and "getting slapped from all sides", to be sure, are in the eyes of the beholder - No doubt your "sides" represent DNC types and mainstream media outlets already in the bag for Kerry.

That same report also showed that 236,000 jobs were added over the past year, bringing the net job loss since President Bush took office to 500,000, and not the 1,000,000 on Terry McCauliffe crib sheets and blaster Emails.


This links to Zogby's summary of their interactive poll method:


The results are weighted by party ID, so even if there are more Dems (which is probably not true), the weighting scheme compensates for that. Interactive polling is still relatively new and has its own unique set of problems, but telephone polling has problems that in many ways are worse. The only difference between the two right now is that telephone is more established in the public mind and therefore more acceptable, but as a practical matter online polling is as accurate or more accurate than telephone.

Zogby is NOT a partisan. It makes no sense for a polling company to undermine its credibility by favoring one party over another. Its reputation is based on its ability to deliver accurate results, not serve the Democratic or Republican party.

Interesting discussion of interactive v. telephone polling. At least the interactive is weighted by party ID, something many of the telephone polls have ignored. I've heard that no cell phones are used in the telephone polls. Doesn't that exclude the large number of us youngsters (& more liberals across the board) who only have a cell phone? Can anyone confirm or deny the assumption?



Phatcat, Regarding:

"Zogby is NOT a partisan. It makes no sense for a polling company to undermine its credibility by favoring one party over another. Its reputation is based on its ability to deliver accurate results, not serve the Democratic or Republican party."

As I've posted on this board previously, Zogby was well regarded as an independent pollster until 9/11; Since then he has increasingly become sympathetic to the views of his Bush Hating brother Jim, the head of an Arab American organ in the US.

We don't know the weighting he uses, so for all we know, he can jimmy the numbers the way he wants them to come out. Moreover, I can assure you that he has many more Dems subscribing to his site than Reps, and that makes his numbers suspect - notwithstanding his claim of fudging, er, weighting, the numbers to balance party ID.

Perhaps most important, He claimed in not so subtle languange around the time of the Abu Gharib prison scandal that the race was Kerry's to lose, thereby undermining his position as an independent pollster in the view of many. Given this previous statement, it's in HIS interest to tweak his polls in order to support his pre-disposition.

But that's OK, both sides have partisan pollsters in this fight: The Reps have Strategic Vision & McLaughlin, the Dems have Zogby, ARG & Demo Corp, among others.

Just because you're a partisan pollster doesn't mean you're wrong - It just means people have to look at your polls through the prism of your bias.


The problem with weighting is knowing the appropriate weights to use, which has been a topic of some discussion around here.

Phone polls don't target cell phones. The argument I've seen is that there are not that many people who just use cell phones, and that they can weight to get appropriate age relevant numbers. But you're left with the question of whether those with just cell phones are typical of others of their age group.

Speaking of partisan.
Good old Bill Schneider and "CNN" were nice today admitting the regime were doing JKs work for him this week, what with the various blunders regarding Iraqi Weapons programmes, Rumsfeld's statements on Saddams' links with ALQ. and lack of troops on the ground (Bremner)

Lastly they even admitted JFK ahead in the elec. college!!! It was something like 245 (JK) to a meagre 218 (for Pres Awol) with 6 'swing' states going either way. Whooh way to go eh...?

It is true that cell phones are not called. This probably has a very little impact in polls, since all polls are weighted by age groups.

It is possible that people who only own cell phones MIGHT BE more liberal then other young people, therefore making that age group seem less liberal then it really is. but this is just an assumption, i have'nt seen any evidence to back this up.