« Why the Race Is Closer Than People Think | Main | Bush Leads in ICR Poll »

Kerry Takes Lead in New National Poll

John Kerry leads George Bush 46-44 percent in a head-to-head match-up among nation-wide RV's, with 10 percent not sure, according to an Investor's Business Daily/Christian Science Monitor/TIPP poll conducted by TechnoMetrica Market Intelligence Sept. 7-12, 2004.

Comments

From the end of the linked article:

"The zeitgeist still points to a Bush victory. Some 48% of those polled foresee Bush retaking the White House, while only 16% think Kerry will win. Another third (31%) feel the race is too close to call."

So about 46% of those polled will vote for Kerry but only 16% actually think he will win?

How many do you suppose would vote for Kerry if they actually believed he had a chance?

Well, that was excitin', wasn't it? But now Bush's little bounce has come to ground. Hope you got your bets down with favorable odds last week.

The poll also notes that among independents Kerry holds a 10 point lead (12 points in battleground states). This is really great; if Democrats just keep the faith and vote, we will win.

Wow 10 point lead amongst independents generates only a two point overal lead? Looks that party ID balance is wrong again.

Also 12 point lead in the battleground states is overwhelming.

Polling is really getting odd.

Added note: how many overseas voters this time? Previously they tended to vote Republican, because they were high up executives, now that could change radically. 5 Million?

Add me to the list of folks thinking "geez, polling is weird these days."

Do we know about how solid this poll is?

And can we take any bets on whether this will get as much play in the national media as the ones showing Bush way ahead?

Do you know what would really get the left going? A poll that showed Kerry and Edwards leading in North Carolina, even if it was inside the margin of error.

I keep waiting for the "Bush Stalled" or the "Kerry Comeback" headline. The effect of the RNC convention should be known by this Saturday, perhaps waiting till even Monday to get all the polling in.
It would be nice to see a random oversampling of Dems that might feed the storyline.
The debates are crucial, and the more Bush is unscripted in front of a camera the better off we are. If the "Bush Surges" storyline dies down, it will be much harder for Bush to avoid debates.

> A poll that showed Kerry and Edwards leading in
> North Carolina, even if it was inside the margin
> of error.

I'm sure that with the right amount of money, that could be engineered. Polls these days seem more like PR campaigns than scientific surveys.

Is this poll traditionally conducted, or is there something unusual about it, such as poll-by-internet or touch-tone? Does it weight for party ID, as per Ruy's earlier post?

Does anyone have a link to the methodology?

Re: The "zeitgeist" comment above. The poster is absolutely right. Forget about arguing issues or even persuading voters. The single most important thing the pro-Kerry blogging world can do is encourage a sense of confidence among our own supporters. When it's OK to support Kerry, Kerry will win.

"And can we take any bets on whether this will get as much play in the national media as the ones showing Bush way ahead?"

Well, on CNN, Blitzer was dedicating a LOT of time to the new WI that showed Bush ahead and I didn't hear anything about the CSM/TIPP poll. I think it should just be taken for granted that cable "news" (CNN, CNBC, Fox, MSNBC) is just an extension of AM hate radio and that it will push the Republican line all the way to the convention and beyond.

Virt

In my last post "convention" should be election.

Virt

Virt-

WI? I'm not sure what that stands for.

Thanks

A problem with polling in general this year is the large number of people who made up their minds back in February about whom to vote for and are unlikely to change their minds now under any likely scenario. When you take 80+% of the electorate out of play then your typical national poll, with a sample depth of 1000 respondents and MoE of +/- 3%, effectively becomes a poll of 200 people with an effective MoE of more like +/- 7%.

It also doesn't help that pretty much everyone, almost always ignores the existence of the MoE completely -- talking about a 2% "lead" in a poll with a 3% MoE as though it meant something. Even professor Tiexera is routinely guilty of this. But I have to admit this is more tempting than usual since on average, by my reckoning, there have been only three, two-week periods all year long when either candidate has opened a lead on the other in excess.

Once in early April, at the peak of Bush's $40+ million smear campaign against Kerry. Once in the wake of the democratic convention when Kerry got his "bounce," however feeble, and now in the wake of the Republican convention. Bush's bounce is actually starting to look every bit as feeble as Kerry's BTW, if you look at the totality of the polls and ignore all the hoopla.

I don't know what to make of this survey, except as one of several to be averaged together. Perhaps it moves the ball a bit in Kerry's direction.

I do know about the Investor's Business Daily, however, and there is no way they would consciously slant their polls to help Kerry. Same for Fox News.

We have to ignore CNN. They only talk about their own polls, and their Gallup polls are ridiculously unreliable. Given their artificial cutoffs (55% nationally and 60% in Wisconsin), they will always overstate Republicans in their samples. And yet they won't discuss their methodology, their results counting all RV's, or other polls that show that they are totally insane (like a 14% Bush lead in Missouri last week). IGNORE CNN!!!

I meant to say:

"...there have been only three, two-week periods all year long when either candidate has opened a lead on the other in excess of the MoE:

- once in early April, at the peak of Bush's $40+ million smear campaign against Kerry.

- once in the wake of the democratic convention when Kerry got his "bounce," however feeble,

- and now, in the wake of the Republican convention.

Bush's bounce is actually starting to look every bit as feeble as Kerry's BTW, if you look at the totality of the polls and ignore all the hoopla."

We need to hear more from Teresa Heinz Kerry because she "Gets It" and knows what it will take and how to WIN !

WASHINGTON - Teresa Heinz Kerry said Monday that President Bush is unwilling to change, a character flaw that shows "inattention and indifference" rather than strength. Of Bush, the Republican-turned-Democrat said the president "demonstrates he cannot and will not change. The president thinks this shows strength when all it shows is inattention and indifference."

THK has cleverly opened a gaping hole in Bush's credibility with her spot on characterization of inattention and indifference as a false sense of Bush strength. Bush sees everything in black and white which shows arrogant stupidity.

This should be much more alarming and concerning to Americans i.e. Bush is inattentive which is dangerous so cannot be trusted.

More importantly THK has served up a distinguishing contrast who said of her husband, "His value in understanding complexity and not being afraid to face it and take action and follow through have been a trademark throughout his life."

I believe more Americans would prefer this type of person in charge as President.

Paul Krugman would agree.

Taking On the Myth
http://www.nytimes.com/2004/09/14/opinion/14krugman.html
September 14, 2004

"The truth is that Mr. Bush, by politicizing the "war on terror," is putting America at risk. And Mr. Kerry has to say that".

"George W. Bush has NOT earned our TRUST, is a FAILURE as President, and should be FIRED".

I also followed the link to the IBD/CSM/TIPP poll, which handily included all the recent national results. Averaging them all together gives Bush a 3.6% lead among RVs and 4.4 among LVs (and we know how reliable LV screens are.) Not insurmountable margins.

His lead narrows to 2.6 among RVs and 3.7 among LVs when I leave out the weird Time poll which has double-digit margins in both samples that are out of whack with the others (I'm also dubious about the quality of the sample of the ABC/WashPost survey but included it.)

Earth to the punditocracy: it ain't over yet.

Mike in MD,

I don't have my spreadsheet here (logging in from work again) but I have noticed no particular tendency in IBD/CSM polls to deviate significantly from the average of other national polls.

-C

You have to be an idiot to not vote because you don't think Kerry will win. What happened to all you Democrats who voted for Kerry due to "electability?" Remember that term?

Yes, I'm a sour Deaniac. But I'm still voting for Kerry for the love of all that is still holy in this world.

>>>
Earth to the punditocracy: it ain't over yet.
>>>

Right?!?!

Even the normally sensible Mark Halprin and Co. seem to have gotten swept up in the general ululation. It is kind of fun to watch them all running about like their heads were on fire though.

Cal, I wasn't saying that IBD polls tended to deviate from others. I was making a concealed point about the conservative tilt of the IBD (especially its editorials) and that if its polls are showing a Democrat even or ahead, then that's probably good news for the Dems because it isn't the desired outcome of the publication, but rather reflects an honest count and weighting of the results.
The new IBD poll showed averages of recent (post-GOP convention) polls and it wasn't out of line with the others. In fact, the IBD survey was one of the more favorable ones for Kerry, but taking margin of error in to account puts it right in the mainstream.

Interestingly, it's the Time survey, by an ALLEGEDLY (emphasis mine) liberal-leaning publication, that seems to be the outlier, on the pro-Bush side.

No offense Adrock, but I believe the term "electability" still very much holds true.

This was always going to be a down and dirty, tooth and nail fight to the finish. Of the choices we had last year, there is simply no question in my mind that we settled on the guy with the single best shot at winning. There is also no question in my mind that we've still got pretty even chances of winning, and that's about as good a shot as we were ever very likely to get.

So buck up. We're just getting to the interesting part now. Kerry and Edwards are very much still in the game and this is usually the point where Kerry actually starts getting his head in the game.

Mike,

I know you weren't. I was actually agreeing with/backing up your assertion in a previous post, not disagreeing with your last one. You were just too fast for me, I'm afraid. My apologies for the mix-up.

Do you know what would really get the left going? A poll that showed Kerry and Edwards leading in North Carolina, even if it was inside the margin of error.

There was an August 23rd Zogby Interactive poll
in N.C. showing Kerry with a 1 point lead

Frankly, I've been following the Zogby and Rasmussen polls most frequently. I know Rasmussen is a Repug, but at least his poll is a rolling continuous average with a sample size of 3,000. His poll today had Bush ahead by 0.6 points.

I find telephone-only polls hard to believe. I know my wife and I have Caller ID and we don't answer the phone after 7 pm unless we recognize the caller. Further, how about all of the folks that only have cellphones these days?

Kerry is not where he wants to be, but he has made strides to get back in this thing and by the time of the debates, he should be dead even with Shrub. The key for Dems will be the same as with every Presidential Election . . . turnout, turnout, and more turnout.

re ruy's earlier post on LVs v. RVs. somebody needs to clue the left-of-center media on the unreliability of using LV numbers this far out. today's salon war room, for example, runs through most of the recent national polls w/ nary a mention of whether the results measure RVs or LVs. and they too seem to come to the erroneous conclusion that bush has a big lead.

are ruy or any other polling experts working to correct this?

With respect to the party balance issue, I wonder why all polls don't routinely take party identification into account, at least in a somewhat minimal, but statistically quite sound fashion.

Namely, what one could ask a respondent is, did they vote in, say, 2000, and were they at that time a registered Republican, Democrat, or neither?

While not all voters will have voted in the year in question, a good number will have (particularly among the RVs and LVs), and for THOSE voters, one will have hard voting data with which to adjust the numbers -- every bit as hard, I'd think, as age.

"WI? I'm not sure what that stands for."

Sorry, WI = Wisconsin. Blitzer was spending a huge amount of time going over a new poll (I think it was Gallup) that showed a large Bush lead there...nothing on the polls showing it a close race anywhere else.

Virt

What is ICR and why do they show Bush with a decent sized lead?

If you follow almost exclusively the campaign news, it seems that Kerry has really stepped up criticism of Bush in the last few days. Bush has been slow to respond to the Assault Weapon expiration debacle, and offered only rambling denials when criticized for the cost of the Iraq mess.

I think Bush's backslide has already begun.

frankly0,
I think the problem might be that respondees tend to say they voted for the eventual winner.

A rolling dynamic weighting might be the best idea.

Really the raw numbers would be the most helpful.

Thanks, Virt.

That was Gallup in Wisconsin, as it happens. 4% difference in the RV, 8% in LV. Imagine that with Gallup

Ray Stewart is absolutely right on the phone polls. I never answer my phone unless I reconize the caller. And also, what wbout the nation wide "do not call list"? I understand there are millions signed up for that. I am one of them and I don't get any sales calls anymore and I used to get them several times a week.

Ed-

I think for the do not call, you can still get survey calls. They just can't sell you anything. I know that I don't get solicitations but have gotten a survey or two since signing up.

Gail,

I don't know much about ICR, but I see that there's the usual RV/LV split. They show Bush up by 7% in a 3-way with LV's, but only 4% with RV's. That 4% is quite consistent with what we've been seeing suggesting there's still a small Bush lead.

Okay, so I don't answer phone call surveys either.

Want to be that somebody has done a poll of people who refuse to answer phone poles?

I suspect that the annoyance factor is bipartisan.

NOT TO WORRY

I want to remind everyone that Mark Mellman, Senior Strategist for Kerry-Edwards Compaign, said in a memo August 24th, 2004 before the Bush convention:

"Incumbents who went on to win reelection had an average lead of 27 points after their convention. Indeed, the average elected incumbent--winners and losers--had a lead of 16 points after their conventions. An average bounce would still leave Bush well below the historical mark set by other incumbents, particulary those who went on to victory.

Perhaps most important, the average elected incumbent experienced a 4-point drop in his share of the two-party vote from the post -convention polling to Election day. Thus, to beat the odds President Bush will need to be garnering 55% of the two-party vote after his convention. Anything less than that and the president will remain in grave political danger.

Here's a link to ICR:

http://www.icrsurvey.com/about_2004.html

First, the bad news. Looking only at RV's, in the August 25-29 sample, they found a 48-45.3% Kerry lead. In the Sept 1-5 sample, they found a 48.2-47.3% Bush lead. In the Sept. 8-12 sample, they found a 47.6-44.1% Bush lead. So the trend has been in the wrong direction, though from Aug 29, it's only Bush gaining 2.3%, Kerry losing 3.9%, both close to or in the margin of error.

Next the good news. They show Bush losing ground from the last survey among independents, down 6%. Of course, with the small sample, that, too, will be within the margin of error.

I'll stick with what I noted before. Seems like there's likely a small Bush lead. Plenty of time to overcome it. If we can get the media to actually focus on substantive stuff the rest of the way.

Ed -

I suppose, then, we shouldn't have to worry, since BC04 has only displayed an "average" amount of dirty politics, and will only spring an "average" October suprise?

I don't want Kerry supporters to be reassured - I want them scared enough about the possiblity of an "average" second term to work their asses off over the next two months to mobilize the vote.

Regarding the Time Poll

Let's start with the observation that the split between Gore voters and Bush voters was roughly 50-50 in 2000. (In fact, Gore had more voters than Bush did.) OK, now what do we do with the latest Time Poll that had Bush at 52% and Kerry at 41%. Well, if we look at the details at www.srbi.com.

http://www.srbi.com/condensed-data-2004-5.pdf

F11. Did you vote in the Presidential Election four years ago, between George W. Bush and Al Gore?

(among 1013 RVs)

Yes 87%
(that makes a sample of ~881 people who claim to have voted)

I presume that this is the base for the next question:

F12: Regardless of how you feel today, did you vote for Bush or Gore four years ago?

Gore 41%, Bush 53%

53% of 881 is approximately 467, let's assume it was rounded up from a lower number. It still must have been at least 463. (Calculations not shown.)

Using a binomial calculator, http://members.aol.com/johnp71/confint.html

I consider the confidence interval around 53% here. If they are saying 463 of 881 sampled were Bush voters, then the 99% interval around 52.5% extends from 48.16% to 56.92%.

But we know that the true proportion of Bush voters from 2000 was less than 48%. (And the true proportion of Gore voters was greater than 48%). How did SRBI take a photo finish election into a 12 point margin?

If this were an unbiased sample, we would only see this many Bush voters by chance with a probability of less than 1%. In fact, if we look at prior polls by srbi, we see similar skewing towards Bush voters.

The real puzzle is why Time would release such a biased poll. I have to conclude either incompetence or corruption is the cause. A rule of thumb is to always assume incompetence as opposed to corruption whenever given a choice, but in this case, the bias is so glaring it's hard to take seriously the idea that nobody at srbi noticed it.

Oh, I have a Ph.D. in mathematics, if that wasn't obvious by now.

It occurs to me that many of the posters here are mathematically adept, so I didn't need to trumpet my qualifications. Hope that didn't offend anybody.

Ed..

I am not so sure that we can use whole lot of historical data on these elections. The conditions are quite different and the candidates are quite different too. I think, as in 2000, these elections will hold their own place in history.

I really do believe that the media will call these elections wrong again and I also believe that the polls are out of sync with the reality which will play itself out in the booths in Nov.

I find very few writers who step right out and support Kerry's positions on any issues. They however laud bush on character and strength. I think Kerry will surprise these writers.

Too many people are being misled by Bush's apparent strength. No one seems to make any effort to tie this strength and character to anything of significance or importance. For all his strength and character, what has he used it to achieve? What are the positive things that comes out of his strength?

For all his character and strength, Bush simply cannot properly deliver on his promises from 2000. He is really a weakling.

The problem is maintaining the difference between Kerry supporters being battle-ready and being discouraged. I've run into any number of fair-weather dems who've already begun the "we're doomed/Dukakis all over again" thing. The media persistence in touting the LV numbers, showing Bush with a substantial lead, has the potential to wreak important psychological damage. I always believed it was the fact a lot of Dems had been hustled into thinking they were going to lose in 2000 that left them ill-prepared to fight in the aftermath.

I see this race as perfectly acceptable for now -- Bush remains under 50% in all RV match-ups (barely above in even the GOP-friendly LV numbers). This is very much like what Carter/Reagan looked like at this point in 1980. I doubt Bush will fall as hard as Carter, since his overall approvals are 5-7 points higher, but he's still very much in incumbent-loss territory.

Perfect kharma is for Bush to win the popular vote and Kerry to win the electoral vote, and therefore the election. I can think of no better scenario.

blamethegreens ,

Absolutuly! we should work our asses off but in confidence not running around like a bunch of scaredycats. When I start loosing confidence I just take a look at the memo frem Mark Mellman and can say to myself that things still look pretty good for Kerry.

How does one get the state-by-state breakdowns from this. We in Ohio need to see this.

"re ruy's earlier post on LVs v. RVs. somebody needs to clue the left-of-center media on the unreliability of using LV numbers this far out. today's salon war room, for example, runs through most of the recent national polls w/ nary a mention of whether the results measure RVs or LVs. and they too seem to come to the erroneous conclusion that bush has a big lead."

I wrote such a letter to Time today, after reading their on-line story with the angle, "Can Kerry come back?" It was easy to click on the reporters' bylines, compose the letter and send it. I STRONGLY ENCOURAGE ALL KERRY/EDWARDS SUPPORTERS ON THIS AND OTHER SITES TO TAKE ACTION AND WRITE THESE KINDS OF LETTERS. It's not enough to whine and wring our hands on these message boards; that accomplishes NOTHING. Take a few minutes and dash off some e-mails or letters.

Sorry to be a downer but Strategic Vision has the Moron-in-Chief up by 10 in Ohio. Don't know if it's LV or RV and I know they're Repub leaning. Also dusturbing - forgot the source - a report has Bushco pulling resources out of Ohio as they consider it safe.

Strategic Vision is not simply Republican leaning, it is a Republican pollster period. That is, SV polls FOR Republicans. Take this poll with a grain of salt as Rasmussen's composite daily polls show a much closer race. Oh and Rasmussen could actually be called Republican-leaning.

Thank you Delilah. Even the LOOK of their website creeps me out!

I once read that local coverage of K/E's campaign stops is extremely favorable. Maybe tapping into some of these local papers would be spirit-boosting.

Rick, your analysis is quite interesting, and it appeals to an underlying fact -- the proportions of people who actually voted in 2000 for each of the two candidates -- that is every bit as good an indicator of whom they might vote in the future as whether they were Republicans or Dems. And, quite importantly, it is a hard fact that the 2000 election was essentially 50-50. While some people might not be honest about whom they voted, it seems unlikely, factoring cancelling dishonesties into account, to affect the numbers much one way or the other.

And a 12 percent difference is, of course, actually LARGER than the recorded gap between Bush and Kerry!

Gail, keep your spirits high. I have my own electoral map which I update for work. I can't afford to be inaccurate. I can't stretch the truth or my credibility would be compromised. And when I look at all the numbers for each state, I think this is still a close race and it could go either way. Now that analysis is based on the numbers and does not consider any new voters or youth voters who may not be counted in the calculations we tend to see when we simply look at the polls. So, keep the faith -- we are still in the game and we still have almost two months to go.

And, Delilah, my sense is that the great majority of new and young voters will vote Democratic.

A I in the minority here? I personally think that this so called jump by Bush after the convention was planned months ago. I think they planned on who and where to call to get the numbers they wanted just like they cherry picked info about Iraq. Figures don't lie but liars figure. Hell, this is what they do. So many of you write like this is the gospel truth. Hell, these thugs are liars, plain and simple.