Wednesday, December 15, 2010

Could Alf Landon Save Canada? I Know. Alf Who?

"Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed." - Herman Melville
The Literary digest was a weekly magazine, that had for decades, predicted the results of the American presidential elections, often within a percentage point.

They did this with mail-in polling.

But in 1936, they couldn't have been more wrong, when they determined that the Republican Governor of Kansas, Alf Landon, would beat out Democratic President Franklin Roosevelt.

And they said that Roosevelt would lose, 56 percent to 44 percent.

But when the votes were tallied, Roosevelt would win with the largest landslide in U.S. history. The Republicans took just two states.

So how did they get it so wrong?

Literary Digest used as their mailing list, the telephone directory and auto registrations. But it was during the Depression when few could afford telephones or cars.

And since traditionally, Republican supporters are more affluent, this meant that they had ignored the poor.

Landon ran on a platform of reducing government spending, at a time when more intervention was required to alleviate the enormous suffering.

Gallup to the Rescue

But one upstart agency, also conducting opinion polls was spot on. They actually predicted the Roosevelt landslide.
Here was a bluff, brash Princeton-based pollster named George Gallup, who had been in business only since 1935 -- and he was predicting FDR in a walk. Gallup called his polling company the American Institute of Public Opinion and billed the Gallup Poll as a reliable index of the voters' mood. Much of the political establishment just laughed at him. After all, his "Institute" was just a tiny office above the Woolworth's on Nassau Street.Gallup, though, had the last laugh. (1)
George Gallup would change the way that polls were conducted, by using the same techniques employed when marketing a product.
Gallup would conduct biweekly polls of a sample of perhaps 2,000 people -- each one chosen, in the time-tested manner of market research, to represent a larger group, including all classes, races and regions. (1)
And the reliability of this method has stood firm for more than seven decades.

But is it now the most reliable?

In the same way that Literary ignored the poor, these polls are conducted using landline phones, at a time when cell phones and the Internet are the most common vehicles of communication.

Some polling firms are conducting on-line polls, but only for those already politically engaged, who seek them out.

This method ignores the opinions of the young and the growing number of tech savvy citizens.

Is it time for a new George Gallup? Someone who really has a hand on the pulse of the nation?

The Problem With Polls

In their book Newsmongers, Mary Anne Comber and Robert S. Mayne, devote a chapter (2) to Opinion Polling, especially during an election campaign.

The polls become the story, and the media base their headlines on the daily data. As a result, they often ignore the platforms and policies of the parties. The things we should be paying attention to.

And since opinion polls are attitudinal, trivial things can create glitches. But once the headline is written, it has the capacity to alter the mood.

They can also determine whether or not a person goes out to vote. If the polls predict that their candidate has an enormous lead, they may feel comfortable enough to stay home. On the other hand, if they are on the bottom, the prospective voter may say why bother?

So in many ways they are undemocratic, because only when the vast majority of the electorate show up on election day, can we get proper representation.

Jean Chretien and Sinclair Stevens, both tried to put a stop to polling during this time, but to no avail. (2)

However, I see problems with polling even outside of an election campaign. The media is already predicting another minority for Stephen Harper, despite the fact that we have not had the benefit of an election. Campaign ads are a huge contributing factor, as are debates, in determining winners and losers. But before any of that has taken place, we are told that they will be useless.

The media has held their own election and we were not invited. Or at least only 2,000 of us were invited.

They have become odds makers. Those gurus who try to determine the outcome of a sporting event, using scientific data, like injuries, personal lives, etc. But does that mean that you don't still want to watch the game?

A lot can happen and often does.

How We Can Defy the Polls

The media and several pollsters have had a few upsets recently. They thought that Stephen Harper would bounce back from the 10 point drop suffered as a result of his second self-serving prorogation, because of his handling of the Haiti Crisis.

But that didn't happen, which says a lot for the character of Canadians, seeing as how we now know, just what his sudden interest in Haiti was really about. We would have looked rather foolish.

And the media got it wrong when they predicted that the Facebook group Canadians Against Proroguing Parliament would not have an effect on Harper's fortunes. Yet we were the ones who caused his drop in the polls.

We also were behind the rallies against the suspending of our democracy.

And we can do that again. We can defy those know-it-alls who believe they have the answers. Who are too lazy to talk to Canadians, but instead talk down to us.

They can no longer ignore the poor, the young, the unemployed. Because even for those not poor, young or unemployed, they care.

We're tired of the affluent having all the say and garnering all the attention. There is the potential for an upset.

Let's make it happen. If you really want to upset a Neocon ... VOTE! And if you want to upset the pollsters .... PROVE THEM WRONG!


1. The poll that took America's pulse, By Jon Blackwell, The Trentonian

2. The Newsmongers: How the Media Distort the Political News, By Mary Anne Comber and Robert S. Mayne, John Deyell Printing, 1986, ISBN: 0-7710-2239-5, Pg. 63-74

1 comment:

  1. Polls are ways to spend money and have something to talk about. They are no more or less accurate than they have previously been. And previously, they have been hit or miss.

    If polls were actual predictors of the future, well, we could predict the future.

    Repost: They took polls for 08 as well. I ask you, on election night, were you absolutely sure that what happened was going to happen? If polls were always right, there would be no such thing as election night nerves, or surprises.

    If I am not mistaken, gamblers take bets on elections. Casinos. So do they often take bets on events with a known outcome?

    If polling really worked, there would be much less loss in many business efforts. But it does not work, not in that regard. As a way of making some spin that can be used to send a message that might help you win, polls work well.

    Note the recent use of nameless 'generic Republican v generic Democrat'. All the candidates are known at this point. Asking generic is a way to avoid the actual questions. These elections are not 'generic' in any way. They are in State, with known candidates. The guy running back east has zip to do with who's running here.

    They were polling away in 08 as well. I knew the deal a day or two out, when Bush Bumper Sticker neighbor told me he could not vote for a ticket with Palin on it, and that he was thinking about not just skipping that line, but voting for Obama, to prevent Palin from having office. Very conservative. But not an idiot. No poll would have asked him about that. And had they asked him on that day about a 'generic choice' he'd have said Republican. The choice was not generic. The real choice was Palin or a Democrat or no vote at all.

    If polls predicted the future, those who take them would be stunningly rich and powerful. Polls tell me more about the writer of the questions than about the people who answer them. The intention of the questions is more important than the surface content of those questions.