I’ve learned of a new way for polls to be rigged with justification (as opposed to just, uh, rigged). The Gallup poll is apparently oversampling Republicans in their polls in a way that is out of line with voter turnout in the last 3 Presidential elections. As a result, they’re getting Bush leads with significant margins while many other polls are seeing an even, too-close-to-call race.
Gallup is arguing that more Republicans will turn out to vote, thus justifying their oversample (40% republican, 33% Democrat, rest Independent). The problem is that more registered Democrats have turned out for the past 3 elections. Unless there is some major shift in party registration or an unusually significant voter turnout drive, this is unlikely to change enough to put the Democratic turnout at 7% below the Republicans. In 2000, it was 39% Democrats, 35% Republicans and 26% Independents. Independents have been the group shrinking in overall turnout from their “heyday” in 1996 with Perot.
My real question is whether this is significant and to whom? Do people change who they’re going to vote for based on one national poll? Does a candidates base get more or less enthusiastic (with donations and support and volunteerism) when he’s ahead or behind in the polls?
My gut feels like being behind helps a candidate when it’s a small margin like today, because catching the other candidate is in sight. It motivates me, for example, to give more when I see a candidate falling behind than it does when they’re ahead by a good bit. I wonder if there are any studies of these type of trends…





September 17th, 2004 at 9:11 pm
although I believe polls are a waste of time… the only true poll is the one where people cast their vote on election day.
BUT, with that said, the media has driven a Kerry friendly message all summer long and now a poll decides to push a Bush friendly message and they get slammed.
The reality is that creating news or slanting the opinion one way or another is wrong but how come when it is done to favor Kerry, no one cries foul? But when one does it for Bush… we got whistles all over the place…
BTW, I am heading over to look at the web site of Bram’s candidate. I always like to see what the third parties candidates are all about.
And yes, you are right… Democrats have turned out in the majority ovcer the past three elections. But they had people they liked and for all the anybody but Bush people that are out there in the USA, they just don’t have the same energy and drive to get behind John Kerry the way they did for Bill and Al… that is why they feel democratic turnout will be down… BUT they have forgotten that while the Dems base is not as mobilized as it once was… the Independents are making a comeback in the anybody but Bush category… but some will vote for Nader and other Independents… so the shrinking base is correct… to the degree you mention in your statement above (or more importantly, refernced by the polling company) … no!
September 17th, 2004 at 10:04 pm
Kerry friendly?!? Were we reading the same SBVT coverage? And what Democrats did you know that *liked* Al? Most I know were holding their noses… Including myself.
To answer Sujal’s questions - I’m afraid that polls sometimes create a “herd” effect - the perception of momentum often leads more people to be willing to back a candidate, because everyone likes a “winner.” I think having a candidate behind in the polls only energizes the base, but it hurts with the undecides/less partisan.
September 17th, 2004 at 10:04 pm
Ugh. That’s “undecideds.”
September 18th, 2004 at 7:53 am
I would have to say… if you do not feel the media from Januray to mid-August was not… and is some respects still is… Kerry friendly, then you haven’t read or watched the news. The media want to elect this guy and are doing anything and everything they can to help him… despite his wife’s often errant comments.
And yes, the DEM BASE was 100% behind AL in 2000… it was the added buzz period of the Clinton years (if you need to find an excuse for it) before the hangover of realizing they had lost.
September 18th, 2004 at 8:02 am
Sujal,
Below is the party identification of the CBS/NYT poll. They randomly sample individuals who claim a party identification. What is the gallup policy? Do they know going into the poll or is it random sampling and they ask on the exit the person answering the phone?
PARTY IDENTIFICATION
In most CBS News Polls, Democrats outnumber Republicans among registered voters. Personal identification with a party can change (temporarily or permanently) with events and voting preferences. In many states voters do not register with a party, and individual identification is a matter of choice. In this poll, where the Republicans hold a significant lead in voter preference and more voters hold negative views about Democratic nominee John Kerry, when voters are asked about their partisan identification at the end of the questionnaire, more identify themselves as Republicans. 36 percent say they are Republican, 32 percent Democrats. The percentage that identifies themselves as Democrats in this poll is lower than it has been in CBS News Polls conducted earlier in the year.
——————————————————————————–
This poll was conducted among a nationwide random sample of 1,287 adults interviewed by telephone September 12-16, 2004. There were 1,088 registered voters. The error due to sampling could be plus or minus three percentage points for results based on all adults and all registered voters.
http://www.cbsnews.com/stories/2004/09/17/opinion/polls/main644205.shtml
BUT TRUST ME when I tell you….polls mean nothing… it is to coin your phrase from an entirely different debate… “a feel good method” to make career politicians feel better about the message or their campaign.
And remember this, intelligent people understand polls and their relevance…. engaged people are engaged because of a certain candidate or against a third candidate… the third group is not swayed by polls or lawn signs… they vote the way their friends, families, or acquaintainces do… or they use the excuse “I was born a XXXX (insert party), I vote for the XXXX”…
and finally, if you are apathetic, you remain apathetic.
September 18th, 2004 at 9:12 am
I like how you blame everything on the media, Doug… I don’t know a media that has been more friendly to the governing administration than what we have right now… but then again, you don’t deal in specifics but rather would prefer to remain the victim of a vast liberal (or Democratic) media conspiracy… you’re a fool, and every time someone screams media bias, they’re a fool. The whole media isn’t biased, unless you’re telling me Fox News and CBS News are biased in the same direction…
Most TV reporting is hollow and newspapers are following in that direction…
Kerry, btw, has gotten some of the roughest media coverage I’ve ever seen… he’s getting the same talking points treatment that Gore got in 2000 and the major news outlets (CNN, MSNBC, FN, ABC) are repeating that faithfully.
Keep feeling like a victim, Doug…
September 18th, 2004 at 9:25 am
Leave it to Atrios to point out the interesting part of the CBS poll, how respondents voted in 2000:
Cause that’s how the election turned out last time, huh? Bush with a decided victory over Gore…
There’s something going on here… either this is a blip (if the sample truly is random, you could randomly oversample), an intentional oversample, or more people truly are identifying as republicans. The data above from the CBS poll makes it more likely that this is a blip.
Second, to go back to the poitn of this post, the Gallup oversample is a matter of their policy… it’s a choice they’ve made that they’re not supporting with data. The post I linked to is by someone who has been talking to Gallup.
September 18th, 2004 at 9:40 am
Here is the PDF of the survey results from the CBS/NYT polls that Atrios pulls the above answers from. The oversample in this case seems unlikely the result of party affiliation changes…
September 18th, 2004 at 2:57 pm
I never doubted the gallup policy… I was only asking…if they knew before hand or if they juts asked as some do….when they random sample.
The truth is that anyone who believes in polls is a “fool”….
and no, sujal, you miss the point. I don’t feel like a victim…