No poll results please!!!

In a democracy polls are relevant. Unfortunately what other people think affects our lives.

Polls discourage democracy by placing the means of gauging the public mood in the hands of unelected, unaccountable organizations. Furthermore, they inhibit the democratic process by discouraging voters from casting their ballots in elections that are "already decided."

Yet we cannot prohibit polls.

Thus, we have only one viable solution. We have a duty, as citizens in a democracy, to lie to pollsters.

I have said this many times, plain old:

Consider the value of poll data the same as you would toilet paper.

Both have very high expectation utility, which declines nearly instantly to zero upon realization.

There's a thread around here somewhere that talks about looking into the bowl after you've done your business...analyzing (now there's a double entendre) polls results is the modern-day equivalent of reading tea leaves or goat entrails in order to predict the future.

There's a bit more scientific verification, but just keep in mind that the only poll that counts is the one done on Election Day.

1. We are in the later part of a transition process from classic modern democracy to viodeocratic postmodern democracy. What matters is not the informed opinion, but the general opinion. The important thing is to seduce, not to convince. Image is more important than words. TV is more important than newspapers.
That's how society evolved. We may not like it, but we have to face it.

2. Measuring public opinion is healthy for society, as long as we know we are given a mirror, in which we can see our collective reflexion. Polls should serve on how to convince people about reasons for policy, and where to work for the betterment of society, not on how to follow the whims of the majorities.

3. The "prophecy" effect of polls has been measured is several societies. It is negligible. So the argument that polls discourage or herd voters has been discarded.

4. Through the history of polling, efforts have been made by different politicians and parties to use pollsters as weapons ("gauging the public mood in the hands of unelected, unaccountable organizations). All these efforts have failed, and all that they have managed is to drive corrupted pollsters out of business.

5. Pollsters do know that a percentage of the population in any given sample lies. This percentage has been measured, is small, usually counterbalances itself and is included in the error probability. Too bad that the media very seldom comments that polls give ranges, within the error probability, and that they are not "predictions".

Interesting comments.

As to people lying to pollsters, I can imagine several sorts of lies and motivation for lying.

1.) People who wish their thoughts to remain private might say something other than their real intent.


2.) Taken off guard, people might say something other than their intent by accident.

3.) The old hippie, "messing with their minds" thing.

I'm curious about this statement, since I've never seen such studies. Do you have any links?


Well, there's corrupt and then there's corrupt, fbaezer. I would imagine that the infamous Literary Digest presidential election poll of 1936 would not be repeated today, but there are other ways of skewing polls short of outright fraud or corruption. For instance, if I asked: "are you in favor of preserving the American family?" I'm sure I could get near-universal agreement. But if I then used those results to support an anti-gay marriage amendment, I'd be guilty of a question-framing fallacy. Yet we see much the same thing going on all the time with polls.


Of course, pollsters can predict the percentage of the population that lies based upon polling data.

Polls are like feelings. They can change in an instant.

Joe, if you think I'm lying about the prophecy effect, it's your problem. I've been in, or near the polling business for almost two decades. I will not provide any links.

Quackpollsters are as old as polls, if not older. The Literary Digest Poll is used as an example of what is not a poll (would be as using our a2k "polls" to prove anything: they are fun, but even non representative of the opinion of the A2K community).

Lots of ignorant politicians try to use quackpolls for propaganda effects, thinking on the (negligible) prophecy effect. It just doesn't work. And sometimes the only ones tricked by their use are the politicians themselves.

There is a lot of debate amongst pollsters about the questionaire, from the wording of the questions to the order in which the questions are asked.
For instance, in pre-electoral polls, there is a debate on whether the questionaire should or should not "warm up" the respondents about their political feelings.

IMO, the biggest problem is the media. Since they are prone to finding "top news", they tend to overdramatize the results of every poll, both about "swings" and about the meanings behind the responses. Since many polls are paid for the media, this leaves pollsters in a difficult situation: they must warn about the misuse of their results, but the clients press for spectacularity.

If I had a half a mind I'd dispute this, maybe.

That was completely uncalled for. I never said, or even implied, that you were lying: I said that I was unaware of such studies and I asked if you could provide a link. Rather than getting into a self-righteous little snit, fbaezer, next time why don't you just state, up front, that you can't substantiate anything you say and suggest that anyone who might want to inquire further should go to hell.

Unless you can substantiate that that was "uncalled for" Joe I suggest you do the same for your "self-righteous little snit".

Go to hell, Craven

He's there, his hell is A2K.

Joe, I may be self-righteous. I may be a snit (I don't know what the word means, honestly). I am not little, though.

I don't think I should be forced to provide links to substantiate any of my sayings. You are implying: "I don't believe you; show me a link or else you are a lia". You have the right to believe I'm wrong. Since I've been in or near the polling business most of my adult life, you may think I lie or rely on lies for a living. I don't care.

Now, if you want some data, here's some data (average for OECD countries):
48% of the population exposed to polls and who can remember polls says that polls actually reinforce their previous decision.
48% of the population exposed to polls and who can remember polls says that polls have no effect on their voting decision.
4% of the population exposed to polls and who can remember polls says that the polls have effect on their voting decision: in about 1.5%, it's a bandwagon effect, voting for who leads the poll; in about 2%, it's an underdog effect, voting for the one who trails the poll; in about 0.5% it's an absenteism effect, not voting.

What's very common is that the losing candidates and their supporters blame supposed poll effects for their failed campaigns.

I was once invited into a prestigious polling organization. As soon as they saw a suggestion of the range of my opinions they dropped me like a hot potato. They didn't want poll members given to non mainstream thinking. I can see where they would not want a result skewed by one person (who in a poll represents thousands of voices). Still, my opinions will always be non mainstream if they are never allowed a dignified voice.

That is as unwarranted as it is insulting. I have no reason to doubt your truthfullness, fbaezer, and I have no reason to think that you would lie about the studies that you mentioned. And I would certainly never call you a "lia."

I asked for the links because I wanted to look at the methodology of the studies. Having been involved in polling most of your adult life, I'm sure you can understand that a poll is only as good as its methodology is sound.

Frankly, I am somewhat suspicious of any conclusions about the effect of polling that is based on another poll, and your outline of the data from these "meta-polls" only confirms my suspicions (n.b.: I am not accusing you of dishonesty, I am simply suggesting that the poll results may not be valid). Without looking at the polls themselves, however, I cannot make any kind of judgment regarding their validity.

And this is a snit.

Okay, okay...I gotta ask.

How do you guys change those links into little blue words like that.

I can't tell you how envious I am that you can do it -- and I can't.

Frank: copy the website's URL (for example, http://www.able2know.com) from the address bar and paste it into your text like so:

{url=http://www.able2know.com}snit{/url}
(replace the {} with square brackets [])

The text enclosed within the bracketed BBCodes will then act as an active link.

To put an url in a one word link, cut or copy the url from the website; click on post reply; click on URL icon which is in the boxes on the top of the reply window. Paste url to the space supplied, making sure you don't have two http:// 's;
enter that, then fill in word or phrase you prefer to the long url in the next window that comes up. Enter that.