(Let's join together and stop the...)Polling Panic! ABC Early Voter Poll 10/26/04 - Bush Leads 51-47
Now, several polls showed a Kerry lead in some of the battleground states; that much is true, just as it is of President Obama.
http://www.dailykos.com/story/2012/10/16/1145340/-Reuters-Obama-3-on-Romney-Massive-Lead-in-Early-Voting
However, as per the above link, a Reuters poll this week showed Obama up 59-31 nationwide among early voters. I would expect the gap to close significantly prior to election day, but not enough to narrow the gap to fewer than 6-7%.
I think this is important. In 2004, despite heavy turnout that seemed to favour Kerry, Bush pulled it out. We can speculate about Ohio shenanigans, but the fact is that Bush won the popular vote; this year, it's inconceivable that Obama could do so without also emerging with the 270 electors required for a second term.
We can shut out any pretence that this is going to end up like 2008, but we new that already; even after the convention, I think the majority of us were aware, deep down, that this was going to be close. If you go back through the year, the few polls sampling Likely Voters showed a tie or a tiny lead either way.
In 2004, the Bush campaign was on top in terms of micro-targeting likely voters. Post debates, most polls (bar the odd outlier) tended to show a tiny lead for either candidate. It was a nail biter, but in retrospect, there were signs that Bush was always going to emerge with a greater number of votes: Scroll back through the polls and August aside, Bush usually had the lead. Sure, Kerry was on top at times, but less frequent support is softer support. It's rare that a candidate who has been behind for most of the cycle emerges as President-elect; 1980 is oft-cited but polls showing Carter up were very much the exception after Reagan wrapped up the nomination.
So, with the Obama team's vaunted ground game, a diminished but still substantial lead among with early voters & the fact that it would require an unprecedented swing in a two-man election (forget 1992) to deliver a definitive Romney victory, am I envisaging a 5% win on November 6th? Nah. It will be close - GOP motivation is obviously through the roof, but I cannot help but look at the abnormally large disparity most polls are showing between the preference of registered and 'likely' voters. Could it outperform the usual two point swing? Yeah, but by an additional percentage point or so, not much more.
I've scrutinised the models, mulled over the polls and unless something sensational happens over the next fortnight - a catastrophic debate performance by either man, a terror attack or a final jobs report anywhere from 'pretty bad' to 'incredibly good', I am very confident that President Obama will win with around 50 of the vote to Mitt Romney's 48; more than enough to ensure that he passes 270.
Here's something to toss around your mind:
On this very day in 2004, NBC's poll showed both candidates tied 48-48 among likely voters. With the pool of registered voters? Bush led by 2%, and went on to win by 1.5%.
Today's NBC poll shows both candidates tied at 47% a piece. Among registered voters? President Obama by 5%. I think various factors (hyper motivation on the part of the GOP, polling station antics) will ensure that Obama's 'victory ceiling' is 3%, but as night turns to day on the morning of 7/11/12, I fully expect the news tickers to inform early risers that the President doesn't need to start packing.
One more thing - Please don't have a dramatic moment if Nickelodeon's 'Kids Pick the President' opt for Romney by a point or two. Obama scraped home in that particular poll by a mere 51-49 in 2008. 
= new reply since forum marked as read
