« No Further Comment Necessary Department | Main | Kerry Ahead by 2% in Ohio »

NYT/CBS News Poll: Kerry, Bush Locked in Dead Heat

John Kerry and George Bush are tied at 46 percent of nation-wide RV's, according to a New York Times/CBS News Poll, conducted 10/14-17. Bush's approval rating is 44 percent.


An eensy bit off-topic, but since you didn't blog on it, it is fresh, and others have expressed concern in the past, I wanted to mention that the new Rutgers poll has New Jersey at 51% Kerry and 38% Bush with a 3.8% MOE.

That's for those who doubted me when I said to ignore those other polls, and that NJ would be solidly Dem come the election. My take is that most of the polls are way off the mark and it will be a big Kerry victory in NJ, and probably in the US generally.

If that approval rating in the CBS poll is reasonably accurate and our 50%"rule" for the incumbent is valid, the Monsieur Bush is in deep trouble.

But.... stories of voter suppression efforts -- what happens "on the ground" -- multiply like rabbits and just may save it for Bush.

We're getting more like some banana republic or Soviet Russia.

I will agree with Zogby to this extent -- if we have (or had!) an honest election, it's a Kerry win.
But we know the Republicans are actively suppressing the vote to a possibly unprecedented degree nationally. Jeb Bush was Governor only of Florida in 2000; with W as prez, we're all Floridians now. But unless something is done using those electronic ballots an honest election means a Kerry win -- in my view there should simply be a national law requiring that EVERY vote have a paper ballot, period. Scans are OK, computers that spit out a card that must be cast are OK, and traditional ballots are OK. Punch cards should be phased out nationally too -- even though they ARE paper ballots.
But an honest election means also Democrats fighting to win. Kerry's performance in the debates was quite adequate -- although against Bush [and for all the protestations about his wisdom, Cheney] I thought it could have been much MORE of a blowout. It's like seeing political heavyweights like Jesse Jackson and Dick Gephardt go up against Ann Coulter and punt (tho not as bad as that).
But Kerry must refrain from Dukakissing. It's the ONLY way the Democrats can lose in the absence of something spectacular, like the widely predicted Osama surprise. Kerry needs to resolve the terrorism issue in ways I've described in previous posts; had he really confronted the flipflop spin effectively, I would agree with Zogby.

It may not sound like it, but I mean this to be a serious question. Is anyone forecasting the percentage of votes that will not be counted? For example if 50% of the voters turn out to be Democrats, will 100% of their votes be counted?

If Bush's approval rating truly is at or about 46%, then he is in deep, deep doodoo. The kind liable to swallow one up.

TPM adds another positive poll indicator,
NBC/WSJ poll tonight has a tie among LIKELY voters. The tidal wave is starting to reach shore.

I, for one, would be delighted to go into election day with the race showing a tie and the polling internals as they are today. Undecideds break more toward the challenger as has been said many times; Democrats have the proponderance of newly registered voters and still have the superior groiund game in battleground states. Tie goes to the challenger.

This sounds interesting. From Josh Marshall's blog ("Talking Points Memo") today:

(October 19, 2004 -- 11:23 AM EDT // link // print)

NBC/WSJ has its new poll out tonight. And I hear it's got welcome news for Kerry. Dead even among likely voters.

-- Josh Marshall

Thought you might enjoy a little look back on Gallup's polls from 2000 (ALL LV):
Note: Gore went on to win the popular vote.

Oct17 Bush 48 Gore 42
Oct18 Bush 49 Gore 39
Oct19 Bush 50 Gore 40
Oct20 Bush 51 Gore 40

Oct21 Bush 50 Gore 41
Oct22 Bush 46 Gore 44
Oct23 Bush 45 Gore 46
Oct24 Bush 48 Gore 43
Oct25 Bush 49 Gore 42
Oct26 Bush 52 Gore 39
Oct27 Bush 49 Gore 42
Oct28 Bush 49 Gore 42
Oct30 Bush 47 Gore 44
Oct31 Bush 48 Gore 43

Nov1 Bush 47 Gore 43
Nov2 Bush 48 Gore 42
Nov3 Bush 47 Gore 43
Nov4 Bush 48 Gore 43
Nov5 Bush 47 Gore 45
Nov6 (last poll before the election)
Bush 48 Gore 46

Bush got 47.9% Gore 48%

Dear EDM:

At my site I've listed email addresses for Sinclair advertisers.


You can cut and paste the addresses into your email browser.

Send them an email and urge them to cease advertising on Sinclair-owned stations. Furthermore, tell them that, for the time being, you intend to boycott their products and services.

I got the list from this site: http://boycottsbg.com/

but the link to "email addresses" was down.

I cut and pasted myself.

Have fun!

S. H. Horton

Looks like the internet and sites like this one are beginning to have an impact. The NYT today includes a story about the discrepancies among polls based on RV/LV methodology.
Some good coverage of all the issues everyone here is familiar with. In particular high voter registration is this years wild card. However, there was still no mention of the problem of weighting for party ID.

This is not a comment but an FYI on the most recent Ohio poll, 10/19/04, here reported on by the Cincinnati Enquirer newspaper:
Poll: Presidential race a dead heat here


Enquirer staff report

The latest Ohio Poll finds the race for president is a statistical dead heat in Ohio.

The poll says 48 percent of likely voters support Sen. John Kerry, while 46 percent support President George W. Bush. Five percent are undecided, while 1 percent support another candidate.

The results are similar to the August Ohio Poll, conducted after the Democratic National Convention. Bush led Kerry by 11 percentage points in the September Ohio Poll.

The poll also found that in the race for U.S. Senate, Sen. George Voinovich leads Eric Fingerhut, 62 percent to 35 percent.

The poll was conducted Oct. 11-17 by the Institute for Policy Research at the University of Cincinnati. A random sample of 757 likely voters was interviewed by telephone. The sampling error is plus or minus 3.6 percentage points.

The one common thread in all of the polls is that
Bush is running between 46 to 48%. These numbers have held true in every poll for the last 10 months or more. It is extremely difficult for an incumbent president to be re-elected with support
under 50%. This has been reported on this website
as well as others many times. The reason for John
Kerry's numbers changing have to do mainly with
Party ID weighting as has also been reported. I would pay attention to Bush's poll numbers and
job approval numbers, as they will be more in line
with his actual vote totals after the election. This translates to a President is serious political trouble,a story that is not getting much play in the media.

i have read about four separate columnists (all bush partisans) writing about bush opening up a 'widening' lead. Do they ONLY look at the Gallup poll or what? Maybe their anticipating that Kerry will do with the soft-on-terrorism Bai spin what he did with the flipflop template. After all, THE AGENDA UBER ALLES

Incumbents almost always poll their approval rating. If Bush's approval rating is just 44%, then I'd say he's in some serious trouble. Factor in the late-deciders going for Kerry and we could be looking at the first majority Democrat win since 1976.

The apparent recent swing toward Kerry is good news, but I expect advantage will swing back to Bush and then to Kerry at least one more time before voting day. It's been that kind of election so far, sigh.

And you just know Bush/Rove will drop some big media bomb in the next few days. High profile arrests or terror alert or major military offensive. Double sigh.

The MSNBC/WSJ will be comforting because the Fox poll is not. Fox does weight by party indirectly because it weights by urban/suburban/rural. Today's poll assumes 36% D, 34% R and 30% I. Not unreasonable. But, they have Bush up 49-42 because Bush leads among Indpendents by 51-33. What does it mean Ruy?

Some more good news, if off-topic.

There's a poll showing the Kentucky Senator race tied. And another showing the Democrat within 7% of Specter in Pennsylvania. Maybe there are a couple other places in which the Democrats can pick up a Senator and regain control at least of that part of Congress. Here's hoping.

a consensus opinion from talking heads is that presidents poll on election day at their approval rating, regardless of challengers numbers.

that means bush would recive 44% of vote today.

after headline on chris matthews, scarbourgh country and cnn blare "bush ahead in gallup" the program's guests reject that conclusion almost unanimously. Even scarbourgh admitted his guest made sense regarding "approavl rating" test and battleground state polls favoring kerry.(thus leaving gallup poll meaningless)

matthews had nearly the same guests drawing the same conclusions where upon matthew blihtely opined to tony blankely that bush was going to win. blankely (washington times) didn't seem qutie as certain. he also answereded all the questions put to ron ssuskind by matthews for suskind.

i ceratiinly hope post election these "commentators" are challenged on their predicions and methodology.

one dumb question:

are we to ignore LV's until elction day or eve? or at what day do LV polls have some relevancy?

on an additonal note , i am invovled in A.C.T . GetOutTheVote effort in a swing state. i can only say that it is a very impressive, effcient organization who i beleive will do a remarkble job on t/o in earlyvoting and election day. readers should be very much encouraged by this on the ground effort.

also recommend you find any GOTV organizaiton in your locale to help on election day t/o. any and all organizations can use you.

I find it unlikely that Bush can be leading among independents by anything near an 18-point margin. While I don't dismiss Fox News polls out of hand--at times, they have been more pro-Kerry than several others--this clearly seems out of whack. Those results among independents are not even close to anything that any other poll (even Gallup) has put forth. I haven't seen the Fox poll on any other site except Fox News itself and some references in blogs; maybe this is why.

old records show that upon election day 2000 the polls were showing that Illinois was going to be close, something within the margin of error, but Gore carried that state by 12%

If Bush's approval rating is 44% and he has 46% of the voting population at this point, then my guess is that about 47% is his ceiling, because he's getting votes from people who don't like the job he's doing, but think that Kerry'll do worse. And this poll, I'm sure, doesn't poll people with cell phones who are younger voters. In a 1936 telephone poll, the people doing the poll predicted FDR's opponent the winner, and people who had phones tended to be Republicans at that time.


State by state polls appear more central than national polls. Bush and Kerry can have 40% leads in Utah and Rhode Island, but 39 of those 40 points do not do anything.

However http://www.davidwissing.com has been kind enough to pile up about ten national polls and note that they fall in two groups of about the same size. One group shows a tie. The other shows Bush up by 5 or so. And, since I read D and R blogs (L blogs are more focussed on doing politics than on polls), I see substantial numbers of people who are D or R arguing, one poll at a time, that the polls giving the wrong answer should be dumped, because a whole family of polls says their side is right, with inadequate attention to the feature that the polls are in two families, and the number of members in the other family is about the same.

Without saying who is right, I am reminded of the early warning systems at Pearl Harbor, which gave signal after signal -- radar, submarine in harbor,...--of the incoming event, and signal after signal was ignored as an isolated and obviously wrong datum.