Republican pollsters John McLaughlin and Jim McLaughlin argue today that the skewed demographics of voters in most polls mean that Romney won’t pull ahead in swing states:
Tonight Mitt Romney could give the most effective debate performance since Ronald Reagan out debated President Carter, or even since Abraham Lincoln showed up Stephen Douglas, and Mitt Romney will still lose the post debate swing state polls. Why? Because the voter samples that the media pollsters will call are already stacked to favor Democrats over Republicans. It’s almost like a 1970′s Olympics where the Soviets have the East Germans as judges and referees. There’s just no way Mitt Romney can win these polls. Let’s look at the recent reality of actual voter registration and data in the most critical swing states and then compare them to the most recent media polls in each state.
For instance, look at the demographics of Florida’s voters in polls:
Florida actually lists party registration for their 11.6 million registered voters. 4.6 million voters, or 40%, are Democrats, and 4.2 million voters, or 36%, are Republicans. When over 8 million Floridians vote this year, it will probably be close to the actual registration.
Florida’s most recent exit poll history shows that the average partisan total for the last 4 statewide elections, including the last 2 presidential elections, is: Democrats 37% and Republicans 38%.
2004 Democrat 37% Republican 41%.
2006 Democrat 36% Republican 39%.
2008 Democrat 37% Republican 34%.
2010 Democrat 36% Republican 36%.
Here are some recent biased Florida media polls (compare the level of Republicans):
9/26 CBS/NYTimes Democrat 36% Republican 27%
9/23 Washington Post Democrat 35% Republican 29%
The most recent CBS/NYTimes and Washington Post polls have Republicans at levels not seen in Florida since the 1960′s. How can Romney win the coming media polls with fewer Republicans represented than when Barry Goldwater ran for President?
Full memo here. They go on to wonder if media outlets are facing intimidation by Democrats: David Axelrod, for instance, let Gallup know he was unhappy about their polling results — and then the Department of Justice went on to sue Gallup over a different matter.
When I talked to Gallup last month, Lydia Saad explained why Gallup doesn’t try to make sure its voter samples have certain percentages or percentage ranges of each party and independents:
Lydia Saad, a senior editor at Gallup, argues that if a pollster tries to control what percentage of those polled belong to the respective parties, the pollster might well be predetermining the results. “In a wave election, let’s say,” Saad explains, “there is a huge shift of voters toward a certain candidate. In that case, you’re going to see that party’s ID go up in the polls. Does that mean you push it back down? Well, no, because you might as well just decide what percentage of the vote each candidate is going to get and don’t even bother polling.”
An example of this is how party ID can change around the time of conventions. Some people, for instance, will call themselves Democrats when they plan to vote for Obama and Republicans when they plan to vote for Romney.
That being said, though, I find it bizarre how many polls I look at have a Democratic advantage greater or equal to the advantage in 2008, and how few have Democrats and Republicans tied (as they were in the 2010 election). Seems like if party ID is fluid, it should work both ways more often.