Generic Ballot tightens to six points in latest ABC/Washington Post poll

[okstate]

51-45

http://www.washingtonpost.com/wp-srv/politics/polls/postpoll_110406.htm

This one is way away from all the other polls out there.

Newsweek has Dems up 16 ......... (trend - Dems +15 - Oct 26)
CBS/NYT found it to be 18 ............ (trend - Dems +14 - Oct 8 )
NBC/WSJ has 15 points ................ (trend - Dems +15 - Oct 16)
CNN has 11 points ........................ (trend - Dems +17 - Oct 22)
Cook/RT has 23 points .................. (trend - Dems +22 - Oct 22)

Over the next two days we should see new polls from the following (previous results in parentheses):

FOX News (D +11) - Oct 25
AP Ipsos (+19) - Oct 25
Zogby (+11) - Oct 23
Hotline (+18) - Oct 23
Gallup (+13) - Oct 22
Pew (+11) - Oct 22

Maybe

Harris? (+12) - Oct 9
Time? (+15) - Oct 4
Battleground? (+8) - Sept 27
LA Times? (+10) - Sept 19

[Eraserhead]

Outlier I think.

[○∙◄☻¥tπ[╪AV┼cVê└]

Not meshing with other results.

[they don't love you like i love you]

1/20

[Democratic Hawk]

When I saw first saw that, I was depressed for a minute or two. Let's see what t'others come up with over the course of today and tomorrow

Dave

[Silent Hunter]

This means a number of things:
a) This is an outlier.
b) The undecideds are breaking for the Republicans.
c) There's a last minute shift.

Hold on to your hats, this will be a ride...

[Sam Spade]

This means a number of things:
a) This is an outlier.
b) The undecideds are breaking for the Republicans.
c) There's a last minute shift.

Hold on to your hats, this will be a ride...

This is a correct answer, though B and C are pretty much the same and it could also be a simple "Republican voters are actually going to vote answer".

[merseysider]

I have a feeling this might be accurate; there are other straws blowing in the wind besides this one, like the rising numbers for Crazy Conrad in Montana, Chafee moving into the lead in RI, and the Sodrel - Hill congressional fight in Indiana.

I can't help feeling that the Dems have peaked too early, and that their triumphalism, and that of some of their supporters in the media, may have done them damage; quite possibly motivating Republican supporters who may otherwise have stayed at home.

I've seen something similar in Britain this year in the local elections; although Labour generally did poorly, we pulled back a few seats in our heartland areas, like Liverpool and Manchester. The constant euphoric rejoicing in both the right-wing and liberal-left media about the mauling Labour was going to get actually helped mobilise some disgruntled voters who realised that their party needed them. I saw some miserable-looking LibDems at the count that night.

I've seen too many elections where the champagne corks were popping before a single vote had been cast, but people were drowning their sorrows on election night.

However, if it is an outlier, it may be helpful to the Dems. I remember the General Election in 1997 here. Tony Blair and Labour were up to 20 points ahead in most polls (and of course went on to win a landslide). However, about a week before the big day, a poll came out which had John Major closing the gap to 5%. It was just plain wrong, but I remember that night we were having to turn volunteers away from our campaign HQ! It was a much needed wake up call.

[Democratic Hawk]

Time? (+15) - Oct 4

Latest from Time:

http://www.time.com/time/nation/article/0,8599,1555024,00.html

Democrats 55% / Republican 40% - Democrats +15 (from realclearpolitics)

Dave

[Sam Spade]

Pew Research Center poll just out.

Democrats lead Republicans in the generic ballot by 4 among likely voters.

http://people-press.org/reports/display.php3?ReportID=295

Generic ballot:
LV
Democrats 47 (50)
Republicans 43 (39)

RV
Democrats 48 (49)
Republicans 40 (38)

[Alcon]

Pew Research Center poll just out.

Democrats lead Republicans in the generic ballot by 4 among likely voters.

http://people-press.org/reports/display.php3?ReportID=295

Generic ballot:
LV
Democrats 47 (50)
Republicans 43 (39)

RV
Democrats 48 (49)
Republicans 40 (38)

So, Sam, what do you think?  Ready to throw in the towel yet?

[Sam Spade]

Pew Research Center poll just out.

Democrats lead Republicans in the generic ballot by 4 among likely voters.

http://people-press.org/reports/display.php3?ReportID=295

Generic ballot:
LV
Democrats 47 (50)
Republicans 43 (39)

RV
Democrats 48 (49)
Republicans 40 (38)

So, Sam, what do you think?  Ready to throw in the towel yet?

I don't know yet.  I just post what I see. 

Rest assured, you'll know what I think about everything on Monday.

[AuH2O]

In 1994 the generic ballots had Dem +5. Now, I'm not saying the GOP is going to do well on Tuesday, but a 4-5 point lead is totally meaningless.

[AuH2O]

1/20

Just for a quick math question-- with the Pew out, is your position now 1/400?

[Alcon]

1/20

Just for a quick math question-- with the Pew out, is your position now 1/400?

I'm no math major, but I think if you want to insult him, you should ask him if his position is now 1/10.

Quoting the 1/20 stat is somewhat inaccurate, because 1/20 is assuming the poll is conducted with an representatively accurate sample.  We simply don't know if that's true, here.

[AuH2O]

Umm, no. First of all, I know what the MoE really means, though most people seem not to. Second, if you assume a 1/20 chance of a faulty poll, then the odds of two such polls is (1/20)*(1/20), which = 1/400.

[NHPolitico]

Pew Research Center poll just out.

Democrats lead Republicans in the generic ballot by 4 among likely voters.

http://people-press.org/reports/display.php3?ReportID=295

Generic ballot:
LV
Democrats 47 (50)
Republicans 43 (39)

RV
Democrats 48 (49)
Republicans 40 (38)

You must be logged in to read this quote.

I'll take it. Thanks, John.

[they don't love you like i love you]

1/20

Just for a quick math question-- with the Pew out, is your position now 1/400?

Well I suspect there's different turnout models being used. The thing with the Pew is that the Dems' lead doubles among registered voters rather than likely voters. That seems a bit too much, and it's very possible this election any Democrat is a "likely voter".

Time and Newsweek are also way different, they may be not as good pollsters, but here they mightt just be using a different turnout model. Which model is correct? We'll have to see.

Something else worth noting though, is the Dem support has largley remained the same, only the GOP support has gone up. I suspect there are two explanations:

1-Republicans upset with the situation before were "undecided" but now decided to go with the Republicans
2-Independents upset with the Republicans who are in safe GOP districts are now finally admitting to voting GOP, finally deciding to reelect their safe incumbent. A great example is MN-03. Kerry got 48% there, but Ramstad is safe and should get over 60%. Obviously almost all the Kerry voters would favor Democrats on a "generic" ballot, and then some and probably a majority of the district, but most will vote to reelect Ramstad anyway, so on this poll they'd now say GOP. Basically the race only comes down to how things break down in a few districts. If both Walz and Wetterling only win by only 1-2 points each for example, it doesn't matter if Ramstad gets 62%, even though among those three districts that means the Republicans won a majority of votes, they still are only getting 1 out of the 3 seats

[AuH2O]

Well you're talking about the main question with regard not only to the generic ballot, but all polls. My suspicion all along is that polls were skewed towards the Democrats because the polls were including lots of people with two qualities 1) opposition to the GOP and 2) a low chance of actually voting.

There might be a poll for instance that calls 1000 people and deems 750 to be likely voters (the number of course depends on the sort of screen used to determine a 'likely voter'). But a large number of even 'likely' voters will not actually vote, with all kinds of variables going into that equation.

There could be a poll with a breakdown like this:

1000 registered voters
750 likely voters
500 actual voters

The 500 non-voting registered voters might be 60-40 dem. The actual voters could be 50-50. The RV number would be 55-45 Dem and LV would be somewhat variable depending on the accuracy of the screen, but probably ~53-48 Dem.

So ultimately the question with the polls is how different the non-voting respondents are to the voting ones. There could be circumstances creating an unusually large gap, which will be clear if the GOP performs better than expected on Tuesday. If not, the gap will have been marginal.

[Deano963]

In 1994 the generic ballots had Dem +5.

Um.....no. The last generic ballot poll before the '94 elections had republicans +8, not Dems +5.

[Filuwaúrdjan]

Could anyone who has numbers for generic ballot stuff from previous years post them please?

[AuH2O]

In 1994 the generic ballots had Dem +5.

Um.....no. The last generic ballot poll before the '94 elections had republicans +8, not Dems +5.

Not according to the WaPo/ABC.

[Deano963]

In 1994 the generic ballots had Dem +5.

Um.....no. The last generic ballot poll before the '94 elections had republicans +8, not Dems +5.

Not according to the WaPo/ABC.

Use your head...if they had a 5-point lead, how would they have lost 52 seats?

[Gabu]

In 1994 the generic ballots had Dem +5.

Um.....no. The last generic ballot poll before the '94 elections had republicans +8, not Dems +5.

Not according to the WaPo/ABC.

Use your head...if they had a 5-point lead, how would they have lost 52 seats?

Isn't it obvious?  The Democrats are going to lose another 52 seats this election cycle.

[AuH2O]

That, or generic ballots aren't too important...