It’s a common phenomenon. Here in Canada, where you have three major choices in a federal election, a lot of people are the same way. In the last election I had people tell me they were CERTAIN the Liberals would win because, why, everyone they knew was voting Liberal. The Liberals suffered their worst election defeat in the history of the country, as was in fact predicted by the polls.
I would think that the market would be factoring that in if that information was available, but the Dow is up 500 points this month. Both the Dow and S&P 500 are up 3% in September.
Because the latter posters are assuming a 2008 type turnout while Gallup and Rasmussen (and ABC) are not.
This doesn’t have any effect on basing your model on a better Democrat turnout than 2008.
No idea, but it doesn’t square with actual voter splits in each state that we have available nor does it make sense to assume higher Democrat (or lower GOP) turnout than in the 2008 elections where-- in the above shown swing states-- the GOP has actually fared better in new registrations and retentions than have Democrats over the same time span.
So your argument in summary is that your polls are right, the other polls are wrong and that voter registration data I gave you is wrong. That’s quite a strong statement to make.
But let’s take a look at your graph. Judging from the graph, GOP identification peaked around mid-February/early-March at 30.0%. On March 11, 2012 the split was 34.6D/30.0R/32.8I; today it is 34.2D/25.6R/34.2I. GOP identification fell by 4.4% while Democratic identification fell by 0.4%. On the other hand, Independent identification rose by 1.4%. Where did 3.4% of the registered voters go? Did they choose to register with a lesser known party? Did they simply fall off the planet? What exactly?
(And funny this is that today a brand new poll was added to your graph, with a more likely split of 34D/31R/33I.)
The polling doesn’t count registration data (I’m pretty sure- please correct me if I’m wrong)- they just ask respondents which party they identify with. So isn’t it possible that some who formerly self-identified as Republican no longer do? And wouldn’t this make sense, considering that the approval rating of the Republican party as a whole has been going down lately?
No, this wouldn’t make sense, unless you’re going to assume that (some) polls are actually more accurate than state registration data. But you actually have no proof of this except you wish it to be so, and only those polls which support the underlying assertion that there is a mass exodus away from the GOP. The better way to go would be to look at state by state registration data, where available, and see which poll most accurately picks up the changes between D/R/I.
RCP is simply an aggregate. It doesn’t analyze polls. I don’t care to guess what the electoral vote is though, as I’ve said a few times now, I think Romney will win in a close election.
That would matter only for likely voter poll results. Again the voter ID difference is found on the all RV data; the expectation of what turnout will be is what moves the data back towards a GOP favored position with the open question being how much and with most pollsters having assumed moreso than in 2008. Your analysis does not comport with the facts.
Try again.
Nope. That they are both right for what they each claim to be measuring. Voter registration does not measure current party identification, to believe it does misunderstands what party ID as polled (a fluid quick to respond creature) and voter registration (not very fluid and a lagging indicator) each are.
You do understand how rolling averages of polls work?
Again, I am open to believing that everyone else other than Gallup is committing some heinous methodological error; you however have not offered any explanation of what that error could be that fits the facts of how polls work. OTOH analyses have offered one that makes sense for way Gallup might be off. So far that argument is more cogent.
I, for one, am NOT looking forward to election night, when OMG rubs our noses in his superior prognosticating abilities.
Mostly because it means we’d have a clueless doofus as President, granted, but still.