Posted on 10/14/2008 11:16:19 AM PDT by Titus Quinctius Cincinnatus
I apologise in advance because this email may be a little longer, but it has some very important information that you need to know if you've been following the election in the MSM (mainstream media) even just a little. Please read this all the way through. If you've followed the news, by now you are aware that John McCain is down in the polls. Way down. Hideously down, so far down that he will never, ever, not in a million years possibly recover from the abyss into which his campaign has slipped. For the past two weeks, poll after national poll has shown Obama leading anywhere from 5-10%. Today's Battleground Poll has Obama up by 13 points. McCain's slipping behind in all the battleground states, and Obama is headed for a massive, Reagan-style landslide victory on Nov. 4. It sure looks hopeless for Senator McCain!
Right?
Wrong. What you are seeing is called "cooking the polls". And here's how it works.
We all know that the MSM - the conglomeration of major TV networks, newspapers, weekly news magazines, etc. - are very, very, very, very, very much in the tank for Obama this year. They're not even making a pretense at objectivity this time around. It's so bad that people from other countries, who basically know nothing about our political system, have commented to me that the media are overtly biased towards Obama and the Democrats. That's how bad it is. Well, the first thing to keep in mind is that pretty much all of the major polling houses - Rasmussen, Gallup, Zogby, Battleground, and so forth - are a part of this MSM conglomeration. Many times, they are headed by Democrats, or by people who at least sympathise with leftist politics, and nearly always they have a financial interest in providing the "right" results and keeping the news entities happy. The polling houses hook up with the media outlets, because the media outlets are the ones most willing to pay for polls to be produced - this generates a steady stream of reportable news every three-four days for the talking heads to analyse at length, which helps to keep interest (and viewership/readership) high. Did you ever notice how many polls include LA Times/CBS/NBC/Fox News/Chicago Tribune/ABC/Time or other news outlets in their titles? That's because of this MSM-polling house dance.
Now, since we know the MSM is in the tank for Obama, and since we know that the polling houses are often led and staffed by people sympathetic to Democrats anywise, and since we know that the polling houses are often hooked up more or less directly with the MSM, it stands to reason that there would be a certain slant to the polls being produced and reported by the MSM, would it not? And indeed there IS a slant - towards Obama and the Democrats.
But guess what? This isn't just my opinion, it's not just a hunch. It's verifiable from the polls themselves.
One lesson we should all learn about the polls we see reported in the MSM is to look at what are called the "internals" of the poll - the reported demographic and partisan breakdowns of the respondants, and the proportional weights being attached to each which are used to calculate the final "result" of the poll. While race and age are important as weighting factors, the one we see being monkeyed with the most in recent polling is the partisan breakdown. Essentially, what a pollster does is "weighs" the data he or she receives from the respondants who answer his or her questions. This weighting depends on what the pollster perceives to be the "right" partisan breakdown for a state or for the national scene. Usually, this just involves getting X number of respondents who are Republicans, Y who are Dems, and Z who are Independents, and using their data.
Seems logical, right? If I'm a pollster, and I want a more accurate poll, if I'm polling a state where the partisan breakdown is 35% D - 33% R - 32% I, then I will try to obtain a data set which mimics that breakdown as closely as possible (and will, at the same time, try to do the same thing with gender, age, and race for the target population). Well, it IS logical, and would work reasonably well - within the margin of error inherent in any statistical sample.
The problem - and this is my point - is that the polls we are seeing coming out with huge Obama leads, do not mirror demographic or partisan reality at all.
Newsbusters.org has an excellent article detailing with this sort of polling trickery, which I'll link to and also excerpt below:
---------------------------------------------------------
In the kitchens of the Associated Press, it's almost as if the wire service asked its chief cook -- er, pollster -- GfK Roper Public Affairs and Media, to do the following:
Of course we don't know if the differences between AP-CfK's Sept. 5-10 and Sept. 27-30 results were created deliberately, but the results sure look suspicious (both polls are available at PDF links found at AP-GfK's home page).
The more recent poll shows Obama with a 7-point lead among likely voters, both with and without leaners; the earlier poll showed McCain with a 5-point lead with leaners, and 4 points without.
Almost all of this 12-point swing (11 points with leaners) is more than likely almost completely due to major differences between the two polls' samples:
---------------------------------------------------------
Blumer then proceeds to point out the huge change in partisan breakdown between the poll taken Sept 5-10, and the one taken Sept. 27-30, and shows the latter poll's internals for party as a picture in his article (click the URL to see). You may access the internals for both polls to see for yourself by clicking the link embedded in the portion I excerpted. Essentially, in the Sept 5-10 poll, the partisan breakdown was 33% D - 31% R, which is pretty accurate - even this year, the Democrats only have a ~2% advantage in voter registration, a trend which follows actual exit polls from 2006 and 2004.
Lo and behold, in the Sept. 27-30 poll, the partisan breakdown suddenly became 40% D - 29% R. What DIDN'T happen is that millions of people suddenly up and switched party affiliations. Rather, what DID happen is that the Associated Press altered the "mix" of voters whose data were included in formulating the poll. Essentially, the Republicans went from a 2% deficit (right) to an 11% deficit (wrong) in partisan affiliation.
The result - predictably - is that (with leaners included), McCain went from winning by 5% to losing by 7%. That's a 12% net change. Now, isn't it interesting how this 12% net change is almost exactly the same amount as the net change in party affiliation difference? Blumer calculates that, if the partisan ratio had been the same in the latter poll, that McCain would have been up 3-4%. Essentially, the AP invented a poll to show Obama way ahead.
Friends, that is the polling equivalent of Mama's home cookin'.
But guess what? This isn't the only poll that's found it's way through a MSM kitchen. I've looked at the internals for a NUMBER of different polls over the past couple of weeks, and they all seem to have the same trend - oversampling of Democrats and undersampling of Republicans. I've seen it so many times in the past week that it's a truism for me now. Let me point out one example from last Friday. Newsweek published a poll on 10 Oct which showed Obama with an 11 point lead (52%-41%). Click on http://www.newsweek.com/id/163337 to see the poll, there will be an Adobe acrobat file with the internals that will pop up automatically. The demographic profile on pg. 19 of the Adobe file is the information of interest. Notice that we see a partisan breakdown of 40% D - 27% R in registered voters who were polled. That's a 13% divide - a lot different from the roughly 2% divide that actually exists nationally.
Folks, this is why we're seeing the slew of terrible polling for McCain for the past two-three weeks. The media are trying to generate momentum for Obama. The MSM is trying to convince voters that the recent economic problems are causing McCain trouble and benefitting Obama, when (if the polls were actually accurate) this is probably not the case.
Just over the past two days, I've seen some pretty ridiculous polling claims by the MSM. They've claimed Obama is up by 15 in Pennsylvania, even though all the evidence on the ground says otherwise. They've claimed Obama is leading in WV, which is patently ridiculous. One poll came out today saying that Obama is leading in North Dakota - a state Bush won by 27% in 2004. A poll yesterday said Obama is leading in Missouri by 8%. Folks, I spend over the first two thirds of my life in Missouri, living in both rural and suburban settings, so I know the state. Obama is NOT ahead by 8% in Missouri.
There is MSM-driven poll cooking going on, folks, and the purpose is to try to demoralise people who would vote for McCain, trying to get us to not come out and vote on Nov. 4. Don't buy into it. The MSM is trying to hoodwink us, trying to make us think that Obama is "inevitable", when he is anything but. Please, get out and vote McCain on Election Day, get your friends and family to come out for McCain, get anyone you can who will vote for McCain to come out, even if you have to drive them yourself. We need to overcome the ACORN vote fraud, the Democrat dead people's vote, and the MSM poll-cooking and win one for the country!
BTW, I know this is a vanity, and I'm not ashamed to admit it.
JSM is not WAY down in the polls.
I get one vote and I will cast it for Palin and McCain.
They will win because they are right.
The new Whitehouse chef will have to be very good at Soul Food.
I know. Read the whole thing, eh?
Thanks for the effort. A vanity with much substance. Mark and bump...
Excellent analysis.
Sorry I’ll go back, I stopped when it said “way down”.
Sorry I’ll go back, I stopped when it said “way down”.
Sorry I’ll go back, I stopped when it said “way down”.
Sorry I’ll go back, I stopped when it said “way down”.
Sorry I’ll go back, I stopped when it said “way down”.
That's okay, sometimes I go a little overboard with the dramatic effect :0
“Do you have all of the national polls and their internals? It would be interesting to see all of the polls normalize to DEM+2 instead of whatever party affiliation they have conjured up.”
The latest national polls are here:
http://www.realclearpolitics.com/epolls/2008/president/us/general_election_mccain_vs_obama-225.html
Zogby weighs to +2% Democrats from what I have seen.
Later today (5 PM EST, I believe) LA Times/Bloomberg will be releasing their latest national poll.
I don't have every last one at my fingertips, sorry. I've looked at the major ones to come out (including some posted by yourself!) and noticed this trend, which a lot of others have noticed as well. The Rasmussen polls you can't get access to unless you're a "Premium Member" (and I'm not willing to shell out any money to Ras, that's for sure), so I can't speak to them specifically, but I include them in my general comments about polling bias because the data I collated from 2004 for polling accuracy against actual vote totals showed Rasmussen having poor accuracy, and the largest bias towards Kerry of any polling house included in my data set. On that basis and on the basis of the trend in similar national polls and the verified demographic oddities found in said polls, I would say that Rasmussen is most likely off in his partisan breakdowns as well.
I'm bookmarking a lot of these polls, it will make for some interesting reading after the election.
Oh yeah, I forgot, but in the Newsbusters article, Blumer does a renormalisation to the same partisan breakdown as in the Sept 5-10 poll (which was D+2 IIRC), and concluded that McCain was actually ahead 3-4% or so.
Nah. Just a little bowl of arugula will do.
Very good post!
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.