Ever thought polls might just be wrong sometimes?  Not everything in politics is about race, as much as Palin and McCain wish it were.  And if race was a factor in the polls/primaries, what about gender?

As for the mere existence of a "reverse Wilder effect" - wishful thinking.  We need to continue acting like this race is tied so we can wake up happy on the Fifth of November.  Then in the wake of our celebrations we can go in and analyze if the polls were too nice or not nice enough to Senator Obama.

That "ever thought" was not meant to be directed at Mr. Greenwald or Lowell, just proposing a general thought to readers.

Sorry if it sounded confrontational, I just realized it might have...

The Democratic primary electorate is not even close to the same as the general election electorate. I would LOVE it if this were true. But I think the results of this study have exactly zero predictive power.  

what if it's in fact a sampling issue? The states where Obama overperformed -- generally southern states, including Virginia -- have substantial African American populations. In the past, blacks have voted at rates slightly below their overall percentage of the voting populace. However, in these primaries ... and also perhaps on Nov. 4th ... it's entirely possible that much of the improvement in Obama's vote totals was due to black turnout greater than their relative demographic levels. [Young people, also, were probably undersampled.]

This is one of Ann Selzer's arguments, and her polling has consistently shown Obama to be doing much better than the rest of the polling world.

We'll find out 11/4.  

side lateral,  up the middle Bradley effect. My alma mater tried that play last week and won...oh...you're talking polls. For Barack/Biden, I hope all those folks who said they will but won't are outnumbered by the ones who said they won't but will (vote for Barack/Biden).

Just need to keep calling, canvassing, lte's etc..  

There was a great piece on Huffpo on The Bradley Effect the other day:

Basically the conditions have changed:

1.  America is less bigoted than it was in 1982.
2.  60 million new voters on the rolls are much less racist than their dead grandparents.
3. Harold Ford - he lost, but he did MUCH better in the actual election than pre-election polling.
4. That famous, recent poll that said 6% of voters would never vote black, is total bullshit:  How many of those are in dead red states TX, MS?  how many would ever vote for a Liberal Democrat?  how many in true blue NY, CA?  that poll is useless because it didn't target swing voters in swing states.
5.  expect huge black turnout, esp in VA (5th - Go Perriello!) and NC.

What this article doesn't mention is the young voters who ONLY have cell phones (and thus are largely underpolled).  This is the secret 2% to 3% nationwide that I believe will buoy Obama's victory to true landslide status. Nate has more on the Reverse Bradley Effect.

I have a gut feeling, based on the crap we are seeing in Prince William County, and with Jeff Frederick's statements lately, that there will be a small Bradley/Wilder effect here in Virginia; not state-wide, but covering enough of the state to have an impact.

Of course, off-setting that I think we have under-polling of African Americans, under-polling of youth voters, under-polling of cell-phone-only voters, and newly registered voters who have not voted in a General Election yet.

In other states the Bradley effect may be gone, but in Virginia I think it remains potent enough to counter any one under-polled segment of the voting population; but not all all four.

Just be glad that more voters are engaged than ever before, and encourage them all to get out the vote.  Voter registration is over in VA; now we must get out the vote on a massive scale.

The VA polls are complicated by the fact that there's nothing predictable about the electorate this year. There are way too many young people engaged, way too many new registrations, way too many African-Americans to think that we could ever reliably predict the outcome. VA is historically a conservative state with a large military population. But which demographic cancels out the other? These polls are trying to take a picture of a moving object. If Obama loses, will it be because of racism, or because this is a red state, plain and simple? God only knows, and that sense of unpredictability goes a long way toward explaining the nervous optimism emanating from VA Democratic circles these days.

it will be negated by whatever remains of the Bradley effect.

I think the net impact is probably a wash.  We saw the reverse Bradley during the primaries because it was a party primary.  The composition of the electorate in a general election is different.

I'm going to be cynical enough to admit there could be a Bradley/Wilder effect and that we should not let the positive numbers in the polls make us complacent.  Better safe than sorry; so, let the polls boost your morale and then get out and work your butt off just to make sure.

Having said that, and I don't mean to be contradictory, but I think this needs to also be pointed out, the great unknown factor is all the newly registered voters.  They are not being polled because they are not yet considered "likely to vote."  Nor are they on anybody's list of those to be polled.  So, nobody knows what effect that unknown group will have.  But key to winning is getting out those newly registered, first time voters.  They truly are the ones now flying under everybody's radar.  And they could be the ones who tip the election.

Survey USA estimates black turnout at around 25% in Georgia, about what it was in 2004, when it's more likely to be close to or more than 30%, which is a lot of Obama votes.