I've mentioned this in a few posts before. The gap between Obama's lead in the Registered Voters (RV) polls and the Likely Voters (LV) polls is what I call the Christmas & Easter voters.
Christmas & Easter voters are like those soft Catholics who only regularly go to church on Christmas and Easter. There are millions of voters (sadly) out there who can be counted to turn out only in Presidential years. In off year elections, they can be fickle. But - as Archie Bunker used to say - in the "big" elections, they will show up.
Unfortunately, the LV screens that many pollsters use are so restrictive that many Christmas & Easter voters immediately get lumped in the RV category. I have read that some of the LV screens are so strict that if you admit you didn't vote in 2010 (more below), you are IMMEDIATELY tossed in the RV category.
More after the squiggle.
In 2008, Voter Turnout was 56.8%. In 2010, it was 37.8%. That's an almost TWENTY PERCENT drop! Will 2012 be as high as 2008? We hope so. But, it is a sure thing that it will be much much higher than 2010. This chart shows the comparisons between Pres. years and off-year:
Voter Turnout over the years
Now, I understand that data shows that LV polls tend to be more accurate than purely RV ones, however, it is foolhardy to just go by LV polls. Particularly, when you hear about abnormally high LV screens. One person who got polled said that they practically had to act like Tom Cruise on Oprah and jump up and down on their couch to get the poll caller to agree they were "enthusiastic" to vote.
RCP still has the race as an exact TIE using LV screens. I think the real number is 1 to 2% in Obama's favor when one factors in the RV's - especially them Christmas & Easter voters.
edit. Replaced xmas with Christmas. NO offense was intended. Christian myself.