« New ARG Polls: Kerry Ahead in OR, WI, Lags in IA | Main | More on the Nader Non-Factor »

Kerry Ahead by 4 and 3 in Pair of New National Polls

John Kerry leads George Bush 49-45 percent of nation-wide RV's, according to an Economist/YouGov Poll conducted 10/25-27.

Kerry leads Bush 49-46 percent of nation-wide LV's in a new Democracy Corps Poll conducted 10/26-27

Comments

Do you have any comments on the poll internals? Are they any more reliable than all the others out there? Or should we simply consider them as part of the poll "noise"?

WORRISOME DEMOCRACY CORPS POLL INTERNALS

There now have been a couple times in a row when I have been dissappointed in some of the internals in the Democracy Corps poll. Here are some of my "least favorite results."

If I am reading the tabs right, about 8% of the sample had already voted. Among that group BUSH had a substantial lead, somewhere around 13% or so. This is bad enough as everything I have been reading (at least in the battleground states) seemed to indicate that it was the Democrats who were voting early.

What makes this result even worse is that this same block of early voters voted for the congressional Democratic Party candidate almost at the same rate as the Republican candidate. I think the Dems were at like -3. This means that Kerry is running about 10% behind the Congressional Democrats in the sample used in the poll. I know that there are a thousand possible rationals why this can be true and not a bad sign for Kerry, but I am uncomfortable with the result.

Also, I think this is the 2nd or 3rd time in a row that more people are getting an increasingly negative feeling about Kerry then about Bush. To clarify what I mean, due I guess to the negative campaigning, voters are becomming more negative about both candidates. However, the rate of increased negative feelings is higher for Kerry then for Bush.

Finally, I didnt like some of the results on personal attributes. Im not referring to the "strong leader" or "better commander in chief" numbers, I expected Bush to have much better numbers there. But I am dissappointed to see Bush do so well on questions like, "he is on our side," "cares about people like me," and "in touch with day to day reality (thats not exactly how the question is phrased but you know what I mean)".

I know that I am "missing the forrest for the trees" in a way as Kerry is up by 3 in the poll. Still, some of that support is soft and I wonder if it will stick in light of some of the Bush attribute numbers.

By the way, Im not one who typically "freaks" at every poll result. Basically, I have been thinking Kerry will win by 3 since the debates. I still do. I am just posting this to see if anyone has some comments.

Reading the Economist poll, it is devastating to Bush. Over 3000 samples! And they endorse Kerry, after endorsing Bush the last time (we were in London then, we are here now, taking care of business.

1. The Economist Poll is so out of line with other polls as to not be credible; at no time did the Economist poll show Bush with over 45% approval or with any lead of over 1 or 2 points- a result which (like Gallup's polls showing 10-15 pt Bush lead) is so far out of line with other polls as to be suspicious.

2. The Democracy Corps poll shows that of people who have already voted (8% of electorate) Bush leads 49-34. This means that Kerry can still be ahead by 2 points or so among people voting on 11-2 and still lose the election.

I am puzzled by the Zogby polling. There are huge differences in the poll results of his "interactive" net based polling and these tracking polls which are so volatile. One example is New Mexico where last week Kerry is shown as up by 11% and is now down several points. I don't believe in those kind of swings in the course of a week. Same with Michigan. His contradictory polling results undermines any sense of conficence in them. Can someone explain this?

> WORRISOME DEMOCRACY CORPS POLL INTERNALS


I am also worried about a rumor on NRO's insufferable "Kerry Spot" website [http://www.nationalreview.com/kerry/kerryspot.asp ]. According to an unnamed "source close to the [Bush] campaign, tuned in to the conversations at the highest levels", "the last few days have seen a huge burst of momentum in their numbers". "The internal polls show a significant lead in Florida (outside margin of error) and Arkansas is out of play, with a Bill Clinton visit or without. As for most of the other big ones - Michigan, Wisconsin, Iowa, Ohio, and Pennsylvania, internal polls show all too close to call."

"Michigan is seriously looking like a pickup - Bush and Cheney could be there four times in the last four days."

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Let's hope the 'wingers are deluding themselves, again. So far, I think the missing Iraqi explosives story has helped Kerry, since the WashTimes story about Russian involvement has been refuted. The public debate during the final week has so far been almost exclusively about "Shrub's" past mistakes. No worried rumors about the Kerry campaign or anything.

MARCU$

Same question here, maybe Ruy can comment the recent Zogby polls. What's up with these? Michigan going red? That would be devastating.

But it seems not only his polls are volatile. New York Daily News quotes him declaring Kerry the winner while Bob Novak writes that Zogby told him Bush is winning. He explains that away with Kerry picking up 5 percent in Tuesday alone.

Too much coffee? (Or should I have some?)

Andy,
On early voting, most states that allow it are Red States. And early voters historically tended Republican. The big exception this year is Florida, where early voting in Broward and Miami - very Democratic - is through the roof. So I wouldn't pay much attention to early voting nationwide when, with the exception of California, early voting is concentrated in the West and South. No New England state except New Hampshire allows early voting.

OK, OK.

The 'already voted' lead for George Bush is a non-issue. Not only do I not trust any sample that small (81 respondents), technically speaking the MOE on those numbers is +/- 11. So, the Bush "lead" breaks down this way:

Bush: 38 to 60
Kerry: 23 to 45

Do you consider that to be a reliable range of voting behavior thus far? Neither do I. Besides, there is no evidence that people who vote early are representative of the population as a whole. Worry about these things at your stomach's peril.

Regarding the YouGov poll, the sample size is higher because its an Internet survey. The Harris Poll also has a large number of respondents for its online poll. I'm really interested to see how the Internet polls do this year, because they've shown Kerry in a much stronger position than the telephone polls. Its an apples-to-oranges comparison to the telephone polls, because the data collection and weighting procedures are very different. YouGov has shown consistent leads for Kerry and well under 47% support for Bush. Harris' online polls have been closer, but Bush support levels have been around 47 as well.

This could be the year of the Internet poll if Kerry wins by the margins that both Harris and Economist are suggesting.