MyDD :: Kerry Isn't Losing

Kerry Isn't Losing

by Chris Bowers, Mon Sep 13, 2004 at 03:50:19 PM EST

I may have grown delusional, but I really believe this now.

I do not have enough internals from recent polls, or enough of an argument about the steady nature of Party ID to prove the title of this article conclusively, but I have managed to track down internals from six of the ten most recent national trial heats (ABC, CBS, Fox, Gallup, Newsweek and Rasmussen). My analysis of these six internals leads me to believe that not only is Kerry not behind in the polls right now, he may actually be slightly ahead.

Here is the breakdown of the national vote by Party ID according to all six of these polls

Democrats
	 Bush	Kerry 
ABC	  12	 88
CBS	  20	 80
Fox	   9	 82
Gallup	   7	 90
News	   7	 87
Rasm	  15.6	 79.6

Republicans
	  Bush	 Kerry
ABC	   94	   6
CBS	   90	  10
Fox	   92	   5
Gallup	   90	   7
News	   93	   4
Rasm	   87.4    9.6

Other
	  Bush	 Kerry
ABC	   50	   50
CBS	   48	   39
Fox	   40	   44
Gallup	   46	   49
News	   39	   45
Rasm	   40.7     46

With leaners removed, here are the mean standings from these six polls according to Party ID:

	   Bush  Kerry
Dem	   11.7  88.3
Rep	   93.0   7.0
Other	   48.6  51.4

The mean number of undecideds in these six polls is 4.7. Now, here are how these Party ID standings and this number of undecideds would translate into current national standings according to four different levels of voter turnout:

	  Bush	  Kerry
2000	  46.6	  46.7
1996	  46.0	  47.3
1992	  48.2	  46.1
92-00	  46.9	  46.7

These numbers don't look all that bad for a challenger. In fact, with numbers like these, Kerry would have to be considered the favorite. Considering flagging turnout among independent / not affiliated voters, it is hard to imagine that the turnout will look more like 1992 than 2000 or 1996, when partisans ruled the day at the ballot box. Further, consider these numbers if undecideds experience a normal 66-34 split in favor of the challenger:

	  Bush	  Kerry
2000	  48.2	  49.8
1996	  47.6	  50.4
1992	  49.8	  49.2
92-00	  48.5	  49.8

In three of four cases, which also happen to be the most likely turnout / undecided break scenarios, Kerry claims enough of a popular vote victory to almost certainly win the Electoral College. And that does not even include mounting evidence that Kerry is stronger in the battleground than he is nationally. Both Fox and Gallup give Kerry a five point lead in "purple states," while Rasmussen gives Kerry a 2-point lead in the purple. Electoral Vote has Kerry back up to 269, ahead of Bush's 233. The Zogby battleground poll taken during the convention still showed an edge for Kerry in the battleground as well.

So am I delusional? Maybe, but I think Kerry's repeated strength among independents and the battleground combined with frequent questionable poll weighting shows that Kerry is at least equal to Bush right now. There remains the possibility that Party ID has shifted in favor of the GOP during the past two weeks, but I remain a skeptic on that front. If someone could track down the Party ID split for the recent AP, ICR, Time and Zogby polls, I would very much appreciate it.

Tags: General 2008 (all tags)

Permalink :: 42 Comments :: Post a Comment

Rasmussen (none / 0)

I have only done that with Rasmussen, but I do believe that it shows Kerry is probably ahead. Bush got a huge bounce in the one day numbers from 9/11 (as I had expected) and it immediately disappeared today when yesterdays numbers were added. That means that without 9/11 Kerry would be slightly ahead. One day's worth of numbers cannot be relied on to prove anything but it is certainly a hopeful sign.

by herodotus on Mon Sep 13, 2004 at 04:47:31 PM EST

Purple state lead (none / 0)

I've seen it in several sources, but I'm not sure what to make of it.

I think a portion of it is that several purple-blue states (particularly Washington, Oregon, and Michigan) have become more blue, while there are no purple-red states that have become more red -- perhaps Ohio or Missouri, depending on the polls, but that's it. This could give Kerry a lead in purple states without any shift in the electoral college.

by niq on Mon Sep 13, 2004 at 04:56:27 PM EST

Re: Kerry Isn't Losing (4.00 / 1)

It's fun to go on google and find daily tracking polls from 2000 and compare them to what you're seeing and hearing now. This link -- http://www.cnn.com/2000/ALLPOLITICS/stories/09/13/tracking.poll/ -- shows Gore with an 8 point lead over Bush on September 13, 2000. Many of you will recall that Gore held a lead in that neighborhood throughout much of September. Even more interesting than the numbers is the CNN Spin. A four-to-five point Gore lead equals "neck and neck." An eight point Gore lead is "modest." A similar Bush lead these days, by contrast, is evidence of a "surge." Just another example of the bar being lower for Bush on everything.

by WSW on Mon Sep 13, 2004 at 05:10:49 PM EST

Re: Kerry Isn't Losing (none / 0)

I believe the lower bar is becuase these polls tend to favor Democrats as the CNN poll illustrates. :)

by Patrick Henry on Mon Sep 13, 2004 at 05:44:38 PM EST
[ Parent ]

Re: Kerry Isn't Losing (none / 0)

hey - i was just looking at that tracking polls from cnn in 2000 the other day.

anyone have any info on zogby, rasmussen, etc from 2000? it would be interesting to do a day by day comparison at this point.

Visit us at TexasKAOS, where we're taking Texas back!

by annatopia on Mon Sep 13, 2004 at 05:52:51 PM EST
[ Parent ]

That EV.com map... (5.00 / 1)

Actually, it really should have Kerry at 270 -- one of the "too close to call" states is Maine, which awards its two congressional-district EVs based on who wins the districts, rather than who wins the state. The state may look like a toss-up at the moment, but Kerry is safe in the more liberal 1st District.

Hillary Clinton or Barack Obama for President! Beat McCain!

by Alex on Mon Sep 13, 2004 at 05:45:00 PM EST

'bout time someone noticed (none / 0)

Despite the (doubtless GOP-pushed) theme that the Kerry campaign is in disarray and Kerry is losing, the polls tell a different story, and so do his actions.

I've actually been quite happy with how Kerry has been taking the fight to Bush for the last couple of weeks. This is the Kerry I wanted to see, and I'm glad it's now showing up in the polls...

by jonweasel on Mon Sep 13, 2004 at 05:46:18 PM EST

POLLS (none / 0)

boy are you are an optimist. rasamussen has missouri firmly for bush, and new jersey down to
a 2 point kerry lead. also he has bush up 1 in pa.if kerry is to win he has to win pa. and nj and win either ohio or florida.i think in the end he will win but at this point i don`t see him leading.one good thing an astrology site www.starlightnews.com says kerry will definitely win based on his astrological chart(for what that`s worth) check out the website sort of interesting but definiely anti-bush.

by EXBUSHVOTER on Mon Sep 13, 2004 at 05:54:19 PM EST

Re: POLLS (none / 0)

Rasamussen, if I recall correctly, is a Republican pollster. Now that doesn't necessarily mean he's biased, but it's something to keep in mind.

However, I also seem to recall that his state polling, in the past, has been quite a bit off from actual election results. For example, there's simply no way Kerry is going to lose New Jersey. It would be like losing California.

by jonweasel on Mon Sep 13, 2004 at 06:00:45 PM EST
[ Parent ]

Re: POLLS (none / 0)

Four points is firmly for Bush? Hard to figure how that works.

by Chris Bowers on Mon Sep 13, 2004 at 06:02:09 PM EST
[ Parent ]

Re: POLLS (none / 0)

Repost of the same link I gave in response to the question of the CSM/IBD poll.

http://tis.goringe.net/pop/bias.html

As you see, Rasmussen is consistently biased toward Republicans.

by PonyFan on Mon Sep 13, 2004 at 07:12:18 PM EST
[ Parent ]

Well, in the current Rasmussen tracking poll... (none / 0)

...Bush is only leading by .8%...and that's factoring in some biased data from calls placed on Sept. 11th. Factor that out, and I bet it shows Kerry ahead. Soooooooo, if that's biased in favor of the Republicans, Kerry is really ahead by a significant amount.

by Geotpf on Mon Sep 13, 2004 at 07:56:55 PM EST
[ Parent ]

New CSM Poll (4.00 / 1)

Check out pollingreport.com a new Christian Science Monitor Poll (not exactly a left leaning organization) was released today. It was tied 47-47 among likely voters, and Kerry was ahead 46-44 among registered voters.

by franklin200 on Mon Sep 13, 2004 at 06:01:26 PM EST

Re: New CSM Poll (none / 0)

I seem to recall that Kerry was down significantly in several past CSM polls, so this is nice to see.

What can you say. "Comeback kid", huh?

by jonweasel on Mon Sep 13, 2004 at 06:03:53 PM EST
[ Parent ]

Polls and Bias (none / 0)

Just because a new organization leans in one direction does not necessarily mean that their polls will lean the same way. The LA Times, in partiuclar, has shown huge pro-Bush results in their job approval numbers. However, the notion that the LA Times is pro-Bush is hard to swallow, to put it one way.

by Chris Bowers on Mon Sep 13, 2004 at 06:03:54 PM EST
[ Parent ]

Well... (none / 0)

my impression, from several of their past polls, is that they tend to produce better numbers for Bush than some other polling outfits. But they're not outlandishly biased, by any means.

by jonweasel on Mon Sep 13, 2004 at 06:10:01 PM EST
[ Parent ]

Re: Polls and Bias (none / 0)

The LA Time is owned by the Chicago Tribune company which endorsed Bush for 2000.

by leschwartz on Mon Sep 13, 2004 at 06:30:50 PM EST
[ Parent ]

Re: New CSM Poll (none / 0)

A good site to check out the polling bias of various organizations is

http://tis.goringe.net/pop/bias.html

As you notice, it has IBD/CSM/TIPP poll pretty high on the "leans Bush" scale.

by PonyFan on Mon Sep 13, 2004 at 07:10:12 PM EST
[ Parent ]

Re: New CSM Poll (none / 0)

That site might be due for a bit of revision, in view of the last Time and Newsweek polls that showed Bush with a grossly inflated convention bounce due to oversampling Republicans. Also, didn't the LA Times have a pro-Bush outlier last month at a time when almost every other poll showed Kerry ahead?
Later,
Alex

Hillary Clinton or Barack Obama for President! Beat McCain!

by Alex on Mon Sep 13, 2004 at 11:14:46 PM EST
[ Parent ]

Kerry is not losing. (none / 0)

Kerry won't lose. He, in fact, is destined to win.

Three times as many Bush voters are voting Kerry than Gore voters are voting Bush. Plus, Nader is weaker, with a significant amount of his support going to Kerry, and nearly none to Bush. Plus, Kerry does better in the "purple" states than in either the red or blue ones. Plus, the Dem GOTV efforts will equal, if not surpass, the GOP's GOTV efforts. Plus, independents are breaking for Kerry. Plus, Bush's negatives are huge. Plus, the actual issues (Iraq, the Economy, Viet-freaking-nam) all break for Kerry.

There is no legal way for Bush to win this election. Let's see how he tries to steal it.

by Geotpf on Mon Sep 13, 2004 at 06:21:55 PM EST

Thats It!! (none / 0)

I couldn't agree more with you Chris and have been arguing this without the detail you employ above to my friends. But, you're empiracal evidence lends even more and greater support to the argument, and importantly made a lightbulb go of in my head - your conclusions backed by evidence must be why the GOP, using the SCLM as their stooges, is fervently pushing the idea that Kerry is behind, that Kerry is a sinking ship, that Kerry's campaign is in dissaray. The reason . . . an attempt to dispirit Democrats and Kerry supporters, into thinking the election of our guy is a lost cause, and thus, to depress our turnout! They have probably done the same analysis as you, and have concluded that their best shot is to depress us, in order to keep us from turning out in sufficient numbers as compared to their base . . .

by kelvinchapman on Mon Sep 13, 2004 at 06:50:23 PM EST

The election (none / 0)

I see the race as a dead heat at the moment and I think Florida is going to the critical state again. The Kerry campaign made a big mistake wasting a lot of time and money in such probably non-winnable states as Colorado, Arizona, West Virginia and Virginia. Right now, they have to concentrate on holding the Upper Midwest and PA and winning Florida. That's doable.

by Mark Fulwiler on Mon Sep 13, 2004 at 07:00:19 PM EST

Re: The election (4.00 / 1)

Arizona and the Virginias may be a waste of time and money, but Colorado looks eminently winnable for Kerry. Zogby's latest has him leading, and, significantly, Bush has not polled over 50% there since polling for this election began. Given the incumbent rule, that makes Kerry the favorite to win the state -- which, combined with NH and all the Gore states, would give him 273 electoral votes and the Presidency, even if the Bush gang manage to steal Florida again.
Later,
Alex

Hillary Clinton or Barack Obama for President! Beat McCain!

by Alex on Mon Sep 13, 2004 at 07:24:34 PM EST
[ Parent ]

Re: The election (none / 0)

Colorado may be the most likely red to blue switch other than New Hampshire. With Salazar on the ballot, turnout among Hispanics will be much higher than normal, and that isn't reflected in the polls.

by feynman on Mon Sep 13, 2004 at 08:01:00 PM EST
[ Parent ]

Re: The election (none / 0)

I haven't seem Zogby's latest, but I'll take your word that Colorado may be competitive, and of course your electoral vote analysis is correct. However, I suspect that if Kerry wins Colorado he won't need it.

by Mark Fulwiler on Mon Sep 13, 2004 at 09:17:33 PM EST
[ Parent ]

Re: The election (4.00 / 1)

Perhaps not, but I wouldn't bet on that. The current scenario at electoral-vote.com shows Colorado as decisive for Kerry, with Florida, Nevada, and Maine too close to call (but see my earlier comment on the latter state, where Kerry can count on winning at least one electoral vote). Sam Wang's model shows the same thing; it currently gives Kerry a 91% probability of winning Colorado, and only 19% in Florida. Given the strong possibility of Republican foul play in Florida, Colorado could be a crucial hole card for us in this election; it's just about the only likely way for Kerry to win without either Florida or Ohio. Also, because it wasn't supposed to be a "battleground state," BushRoveCo are less likely to have laid the groundwork for stealing the election there than they are in places like Florida, Ohio, and Missouri.

I can imagine a quite delightful scenario where they successfully steal Florida, lose Colorado -- and then get nailed by the Kerry Administration's Justice Department for election fraud, after not covering their tracks sufficiently. After all, they were counting on having at least four years to bury all the evidence.
Later,
Alex

Hillary Clinton or Barack Obama for President! Beat McCain!

by Alex on Tue Sep 14, 2004 at 12:53:04 AM EST
[ Parent ]

Re: The election (none / 0)

If Kerry loses Florida , he'll have to hold unto ~all~ of the Gore states and win New Hampshire and Colorado. But it seems somewhat unlikely that, in a close race, Bush will not flip at least one Gore state like Wisconsin.

I won't speculate on possible voter fraud.

by Mark Fulwiler on Tue Sep 14, 2004 at 01:38:46 PM EST
[ Parent ]

Re: The election (none / 0)

Don't forget that there's an initiative on the ballot in Colorado that would divide up Colorado's EVs. So if Kerry wins, and this measure passes, then Kerry won't get all of Colorado's EVs. He'll split them with Bush. So Colorado voters have to make a calculation -- if Kerry is up going into election day, then the safe bet is to vote against the measure. But if Kerry is down, then the best bet is to vote for it. But given the unreliability of polls and the closeness of the race in Colorado, a person could end up voting against the measure thinking Kerry will win the state only to find that Bush won it.

by smartliberal on Wed Sep 15, 2004 at 12:08:28 AM EST
[ Parent ]

Re: Kerry Isn't Losing (4.00 / 1)

Kerry up in the IBD poll today:

Investor's Business Daily/Christian Science Monitor/TIPP    
poll conducted by TechnoMetrica Market Intelligence. 
Sept. 7-12, 2004. N=899 registered voters nationwide 
(MoE ± 3.5); 674 likely voters.
	
				
			   Bush    Kerry    Not Sure			

Among likely voters:
9/7-12/04		   47	   47	    6			
				
Among registered voters:
9/7-12/04		   44	   46	    10

by Jerome Armstrong on Mon Sep 13, 2004 at 07:04:09 PM EST

Re: Kerry Isn't Losing (none / 0)

And that's even with Nader included:
http://www.pollingreport.com/wh04gen.htm

by Jerome Armstrong on Mon Sep 13, 2004 at 07:04:55 PM EST
[ Parent ]

Re: Kerry Isn't Losing (none / 0)

Can someone explain to me weighting of polls. It seems to me if you weight a random poll it isn't random. It also seems like the pollsters some how know what the results should look like. It probably is a stupid question, but its been nagging me

by Patrick Henry on Mon Sep 13, 2004 at 07:10:03 PM EST

Re: Kerry Isn't Losing (none / 0)

Read back over the last four days of front page entries, it's covered pretty extensively.

by Jerome Armstrong on Mon Sep 13, 2004 at 07:20:29 PM EST
[ Parent ]

Man... (none / 0)

Boy, is it ever!

Invest in nature

by NCDem on Mon Sep 13, 2004 at 07:34:41 PM EST

Re: Kerry Isn't Losing (none / 0)

An important question regarding Chris' analysis is: why are the polls questioning a higher fraction of Republicans than the national average? Could it simply be cell-phones? The elderly and poor don't have as many cell-phones, and this would presumably bias the sample in precisely the way that is observed.......there weren't many cell-phones in 2000.

by feynman on Mon Sep 13, 2004 at 07:47:18 PM EST

This is quite possible (none / 0)

by Geotpf on Mon Sep 13, 2004 at 07:58:23 PM EST
[ Parent ]

Re: Kerry Isn't Losing (none / 0)

Actually, cell phones may be a factor, but probably not in the way you think. Pollsters generally don't call cell phones; there are laws against making any kind of soliciting call to cell phones, because the recipient frequently has to pay for the minutes used on the call. What that means is that pollsters aren't reaching people who ONLY have a cell phone, no land line. These people are mostly low-income (some cell plans these days are as cheap or cheaper than local phone service, and more convenient because of the mobility) and/or young (I'm a grad student; two of my friends, both very liberal, are cell-only users).

Speaking of the young, there's another group that pollsters miss: college and university students who live on campus where dorm room phone numbers are extensions within the university phone system. My fiancee is in that situation; you have to dial the University's number, wait for it to pick up, then dial her five-digit extension. Of course, she's at a school in Canada, so she wouldn't be polled anyway -- but she is a U.S. citizen, and will be casting an absentee ballot in Arizona, where it might actually make some difference. :-)

Hillary Clinton or Barack Obama for President! Beat McCain!

by Alex on Tue Sep 14, 2004 at 01:11:03 AM EST
[ Parent ]

I have a question (none / 0)

During the RNC convention coverage Joe Trippi said somthing on "After Hours" that seems to go against the undecided rule. He said that the 2 to 1 break for the challenger isn't true in a presidential election. The rule is actually reversed. I hope that's not really true. Anyone?

by moses on Mon Sep 13, 2004 at 08:03:53 PM EST

Re: I have a question (none / 0)

It's more nuanced then that, I found from talking with him. I haven't seen a detailed analysis of the matter; but what he referred to goes along the lines that when it's a close election, and a troubled time, the undecideds break toward the known incumbent, and that when it's not that close, or fairly peaceful, the undecideds break toward the new challenger.

by Jerome Armstrong on Mon Sep 13, 2004 at 08:54:12 PM EST
[ Parent ]

polls (none / 0)

Chris,

Given that there is a media bias toward pushing the notion that Bush is ahead, your analysis might even be a bit on the conservative side.
Thanks for posting- we need to spread the word and change public perception....

by global yokel on Mon Sep 13, 2004 at 08:20:10 PM EST

Question (none / 0)

Just a quick question: why are you removing the leaners? And can you post the #s for the leaners? As it is, the mean for Kerry according to the six polls you cite is NOT 88.3 among Dems. It's 84.43. And the mean for Kerry among other is NOT 51.4. It's 45.5. These are BIG differences. Look, I want to believe Kerry's up as much as the next guy, but if we're going to second-guess the polls, we've got to lay out every step in our methodological analysis. Otherwise, it's just a numbers game.

by smartliberal on Tue Sep 14, 2004 at 12:05:30 AM EST

Re: Question (none / 0)

It was the only way to combine them. Some polls had already removed the leaners, and some had not. Thus, I simply combined the Bush / Kerry ratios using the central mean. Combining the ratios prevents you from combining apples and oranges.

by Chris Bowers on Tue Sep 14, 2004 at 02:34:22 AM EST
[ Parent ]

Matt Lauer and Kitty Kelley (5.00 / 0)

This morning I watched Kitty Kelley's second day of interviews on NBC's Today Show. Matt Lauer is so terrible in his interviewing techniques. At one point Kitty Kelley said that Matt Lauer has played golf with George Bush Sr. in which Lauer stated "I've never played golf with George Bush." I went to my computer and googled "matt lauer bush golf" and lo and behold look what came up:

http://www.fieldturf.com/index.cfm?pageView=readFeature&featureID=11

by drplaud on Tue Sep 14, 2004 at 10:29:06 AM EST

You are not logged in.

In order to post a comment, you must be logged in. If you have a member account, please log in to comment.

If not, you can make an account right here. It's quick and free.