Some good news for Romney-Ryan: the trend is your friend

From Doug Ross @ Journal

... there are several reasons I think the polls are fundamentally misstating support for Barack Obama's obviously failed presidency:

1. The return of "The Bradley Effect": Barack Obama, a thoroughly likable and eloquent politician, marketed himself as a centrist and post-partisan figure in 2008. Unknown and un-vetted that year, he could get away with it. Pollsters are likely not accounting for the fact that Obama is a known quantity this year -- and the well-known Bradley Effect could very well resurface.

2. The revolt of the union rank-and-file: If you're a union member aware of Obama's destructive agenda -- killing the coal industry, oil drilling, trucking, and refining (to name but a few), you're not going to respond to some random call with the truth. Union members have to be cautious and they are used to being pressured and intimidated. I can assure you that pollsters are not taking the Reagan Democrat factor into account, just as they missed it completely in 1980.

3. Incorrect sampling of the population: As HillBuzz explains, polling firms are generally using 2008 party breakdowns when the entire country has changed dramatically:

1. Look at the partisan breakdown of the poll and make sure it is the 2012 numbers. The partisan breakdown I am using for the 2012 elections is: D=33.3%, R=37.6%, I=29.2%. We are at the highest percentage of Republican registrations and the lowest percent of Democrat registrations in history. The polls you see that show Obama leading in states like Ohio are deceiving you because they are running Democrat percentages that are 44% or MORE; that’s just ridiculous because the highest Democrat registration advantage in YEARS was in 2008, where Democrats were at 41.4% to just 33.8% for Republicans and 24.7% for Independents. 2008 was the Democrats’ high water mark...

2. Look at the gender gap in the poll and see if it is realistic. The second easiest way to cook a poll is to include a much higher percentage of women than men because women tend to vote Democrat more than men...

3. Look for the racial breakdown of the poll to check if that’s being cooked for Democrats’ benefit. This is similar to the above, but is a little more complicated because on average blacks are 12% of the population but have a very low turnout in typical election years. 2008 was an anomaly because race-pride drove a lot of blacks to the polls... There’s no way that blacks or any other Democrat identity bloc will vote in 2008 percentages, though, so any poll that’s saying these numbers will be the same (or higher!) is bunk...