Are Universities Playing Politics With Polls?

Strange - two very different polls out today with university connections and they appear to be way outside of the norms. My thinking isn't so much that they are trying to influence the election, but I am wondering if they haven't bought the most positive spin from the Obama camp and worked it into the sampling for their polls.

Montana going for Obama? Don't bet on that.

Obama is at 44 percent and John McCain is at 40 percent in the Montana State University-Billings poll released Thursday. The poll reverses others from earlier in the fall that showed McCain with a lead in the state.

And a Big Ten poll I mentioned earlier that had party ID way off from what one might expect. It also has Obama way up in most battleground states.

Just one observation from an early look. In the internals for Ohio the poll sample was comprised as follows: Dem 40% Rep 27% Ind 28%.

In 2004 Ohio exit poll results Party ID turned out to be Rep 40% Dem 35% Ind 25%.

Has Ohio changed that much in four years? We'll know on November 4th.

Yet, also today, both Gallup and the Battleground poll show McCain tightening it up a bit.

Comments

[chris]

"The poll of 403 Montanans was conducted from Thursday though Monday and has a margin of error of plus or minus 5 percentage points."

That's almost 80% of likely voters. Sounds like pretty tight numbers to me.

[hdtv]

The bigger question perhaps is "who cares about Montana?"

[seekeronos]

"--- The bigger question perhaps is "who cares about Montana?" ---"

Montanans care.

And they'd be among the few states, if I recall correctly, that would be able to secede from the Union and have a self-sufficient economy, should the need ever arise.

[jharp]

chris,

You are right that the Montana polls are within the margin of error.

And I'm sure this is good news for republicans.

And Dan, you obviously don't understand how serious these universities are in the undertaking of polls.

I'm not saying they are experts but to accuse them of playing politics with no evidence is ridiculous.

You've been listening to Rush, Hannity, and Fox News for far too long.

[ET]

More Obama deceit exposed:

http://www.ibdeditorials.com/IBDArticles.aspx?id=309565448930564

Hold on to your wallets, successfull business owners. We will soon be supporting the malingerers like jaharp, chris, mj and the boob.

What a proud time to be a democrat (especially if you are one of life's losers).

[chris]

Hahaha. IBD editorials. They're as nutty as that Moonie paper in Washington. You should really get off your highly successful ass and get to work. You've got some funds to redistribute. Cmon, move it!

[ET]

Take your thumb out of you ass, chris. You'll want to have both hands available when uncle barry pulls his truck up to your trailer.

[skeptical]

Dude, read their literature. They tell how they did their sampling. They haven't "bought" anybody's spin. They have a method, look it up.

And I think 403 may be 80% of the total number of likely voters with regular telephones they have up there.

It shouldn't even be close. Idaho is more than 30 points. Wyoming is about 25 points. How is it possible that it would even be close? Couldn't possibly have anything to do with the candidates or their positions.

[chris]

Ooh - my trailer. You sound, umm, bitter. No doubt that you're a HUGE success though. Huge. Really. Big success.

[mark l.]

the number I'm try to grasp is white voters-

cnn had them at 77% of the electorate, going 59-41 in favor of bush, based on their exit polls.

gallup has them at 48-44 mccain.

consider 18% difference in 04, among 121 million voters * 77%...

net difference of 16.7 million for bush among white voters.

plug in a 4% (from gallup) difference for 08, at 121 million * 77%...

net difference is 3.7 million voters.

switch 13 million voters in 04, and the final tally would produce 49 million for bush, 72 million for kerry and you would get a final result that would be kerry won 60-40...

obama doesn't have a 20 pt lead though, despite the closeness among white voters, he has a 5-6 pt lead?!?!

[mark l.]

the only way a substantial improvement among white voters, relative to kerry, would not manifest itself in the polls for obama is if the weighting of white voters is substantially lower.

[mark l.]

http://www.cnn.com/ELECTION/2004/pages/results/states/US/P/00/epolls.0.html

since I'm referencing the above...

while I'm far more interested in state polling, a discussion of net diffrence is still worth having...

2004
net differences by race.

bush +16. 7 million among white voter(93.2 million).
kerry +10.23 million among african americans(13.3 million)
kerry +.87 million among hispanics(9.68 million)
kerry + .29 among asians(2.42 million)
kerry +.340 among other(2.42 million)

These numbers are ballpark, not gospel, as the cnn poll could have been flawed, I didn't account for nader and basically, bush won by 3 million, while the above difference is 5 million. although the difference between 3/121 and 5/121 is about a vairance of 2.5-4%, let the number hold on the larger macro trend.

i'm curious, will white voters increase this year, or decrease?
I'm sure african american will turnout in droves, but at best they will produce 2-3 million new voters to add to the net number.

This may shed some light on my obsession with the demographics, beyond poor weighting for party affilation, given the presses histrical inability to predict it. Race is a far more accurate indicator.

[mark l.]

guessing the turnout is the essense of finding the winner...
NO POLLSTER predicted it in 04, as it was +22% for bush, and +16% for kerry.

Tales of high registration? the only number that could account for massive amounts is from white voters, as they are the largest group by far. Even gallup has mccain beating obama among white voters, so the more turnout from a group that favors mccain, the greater the disadvantage for obama. The scenario of a white voter, having never voted, deciding to go out and vote FOR Obama as their first vote is just not practical, especially the older the voter.

If turnout goes the same as 04, bo will need 44% of the white vote, where kerry drew 41-not accounting for minority vote increase. I can't see him beating Kerry's number, but we'll see...

[mark l.]

guessing the turnout is the essense of finding the winner...
NO POLLSTER predicted it in 04, as it was +22% for bush, and +16% for kerry.

Tales of high registration? the only number that could account for massive amounts is from white voters, as they are the largest group by far. Even gallup has mccain beating obama among white voters, so the more turnout from a group that favors mccain, the greater the disadvantage for obama. The scenario of a white voter, having never voted, deciding to go out and vote FOR Obama as their first vote is just not practical, especially the older the voter.

If turnout goes the same as 04, bo will need 44% of the white vote, where kerry drew 41-not accounting for minority vote increase. I can't see him beating Kerry's number, but we'll see...

two scenarios-if vote is record level, mccain wins in a landslide, with record level being a 10% increase.

vote total below or near 2004, and obama wins.

anything between 0 and 10% is up for grabs.

[jaime]

Why do you people think its still 2004?

[mark l.]

if i were a dem,

i'd be worried about the ibd and ap poll...

the ones that have been tracking, have been repeatedly asking the same people-rasmussen, gallup, zogby...

the number that are undecided, if accurate is a now very likely a trend. 2nd thought voters. The tracking polls in asking a question repeatedly, cement a decision in the subjects mind and have them reaffirm it on a weekly or daily basis.

If they are going off of a group that they knew would respond, they have diminshed their poll to a subset of our culture.

landslide mccain.

[mark l.]

"Why do you people think its still 2004?"

hmmm...exit polls, which once were gold, missed the election by 7-10 pts in favor of kerry, dems believed they were true. reality hits them at 8pm eastern, and they slink off into the dustbin of history. In the end W is reelected.

I tend to see a manifest problem in the polling, dems believe the polls are true.

What is it they say about history repeating itself?

[MarkJ]

Chris,

"You've got some funds to redistribute. Cmon, move it!"

Cute, but you'd better hope there aren't any folks in your neighborhood who'll be eager to "redistribute" their full-metal jacketed rounds if things go south after the election.