« Latest LA Times Poll: Kerry/Edwards Leading By 1% in Horserace | Main | Blacks and the 2004 Election »

Kerry Ahead by 1 in New Harris Poll

John Kerry leads George Bush 48-47 percent of nation-wide LV's, with 1 percent for Nader, 1 percent other and 2 percent not sure, according to a Harris Poll conducted 10/21-25.

Note that, since Harris is using their more restrictive definition of likely voters in this poll (registered, absolutely certain to vote, voted in 2000 if old enough to vote), this result actually represents an 9 point swing in Kerry's favor since their mid-October poll, when Kerry was behind 51-43 among this particular flavor of likely voters.

Comments

Is it wishful thinking, or am I detecting a trend here? Kerry is up in the following: Democracy Corps 2 pts, LA Times up 1 pt, Harris 1 pt, and ABC News/Washingto Post 2 pts. Gallop has Bush ahead by only 4 which, if their internals have the same problems, is probably a dead heat. I suspect (and hope) that Kerry is doing his famous closing, and I think the 380 tonnes of missing explosives will galvanize the fence-sitter into KE direction.

Harris is underrepresenting new voters. Their likely voter screen:

"Likely Voters are defined as adults who are registered to vote and say they are "absolutely certain" to vote and, if they were old enough, voted in 2000."

So if you didn't vote in 2000 (unregistered, unmotivated) you don't get counted. I think this poll is even better for Kerry than it seems on the surface.

Harris seems to be something that slips under the radar, but from what I've heard they're pretty good.

a trend seems to be emerging, and it's not good news for George Bush.

Notice how the WSJ headline writers just can't bring themselves to say that it's Kerry who is ahead in the latest Harris poll. You have to read to the end of the second paragraph to get that info. Compare that with some recent headlines characterizing Bush's 1-pt leads as "surges" or "grabs". I know, it's no called the SCLM for nothing.

The Harris Poll is the most encouraging poll yet. Last week they ran the poll and offered two different LV scenarios - one with those who will definitely vote AND voted in 2000, and the other with those who will definitely vote but did NOT vote in 2000 (and were old enough). In the second scenario, without 2000 voters, Kerry trailed Bush by 2 (48-46). In the first scenario that excludes new voters Kerry trailed by 8 (51-43). However, in the Harris Poll just released, the LV model resembles the first scenario of the last poll. In other words, Kerry didn't just increase from -2 to +1, he moved from -8 to +1! That is incredible, if it is true, and shows a huge closing among undecideds for Kerry. I can only imagine what the number is for the second scenario voters now.

With such a robust sample, you can see that Bush's MoE puts his ceiling at 49%. His ceiling, mind you.

It does seem that Kerry is breaking a hair ahead in national horse races.

I think the BUSH trend is "YOU"RE FIRED." I was out cnvassing for MOVE ON last night. I called in TEN strong Kerry Edwards voters for ten stops. These were all voters that were not likely to be sampled by any major polling system. There is a vast undercurrent of anger towards the current administration which I think is starting to head towards a crest.

Isn't this LV definition the same one that gave Bush a 51-43 edge in the October 14-17 Harris Poll?

Whether this reflects a real shift or just the LV definition working better closer to the election is another question.

You wonder whether Bush in fact has superfluous solid red support, and whether the non-solid-red polls, which really matter, might show 2% more for Kerry.

chris matthews said today kerry had the wind at his back and if the wind is still blowing in kerry's direction tomorow(thursday) kerry will win.

chris matthew called the election for bush exactly 7 days ago.

larry o'donnel spent 10 solid minutes calling swift boat john o'neil a "liar" on scarbourgh 2 nights ago.
not "misleading' ...he called him a "lair" about 37 times , and very loudly. hats off to o'donnel.

This is their online poll, and it should be noted that this results differs markedly from the telephone poll released just a few days ago. In the poll, using the same definition of likely voters, Bush was up 8 points. (With the less stringest definition, Bush was up 2 points). This is a big swing of 9 points, but one was telephone, one was internet.

Personally, I have more faith in their Internet polls.

based on all the polls and interperetations of the polls and increase in voter reg and GOTV efforts and %blacks for kerry and % cuban americans for bush....i was just wondering when someone will tell me how much kerry will win by.

frankly i am exhasuted and would like to take a long rest....any convincing predictions out there?

Well, you can try some of the meta-analysis sites out there. These analyze a bunch of polls through various means and try to estimate the outcome. I think you'll be pleased with the results:

http://synapse.princeton.edu/~sam/pollcalc.html

http://binomial.csuhayward.edu/WeeklyStatus.html

http://arrowheadengineering.com/

http://www.econ.umn.edu/~amoro/Research/presprobs.html

Kerry will win by three. Now feel free to rest. :)