The last minute polls on election day: message from the medium

Philosopher Marshall McLuhan once warned that the medium was becoming the message. This prediction may have come true during the recent Chicago mayoral primary election.

A couple of TV stations indulged in last minute polls, predicting a sweeping victory for incumbent Harold Washington. Then viewers had to spend a long evening watching Washington's prophesized landslide — and the media's credibility — slowly disappear. Washington eventually won, but a swing of 40,000 votes might have resulted in another "Dewey Defeats Truman" photograph.

The camps of both candidates screamed foul and with some justice. Byrne's people felt that their faithful had given up when the polls said she had no chance. Meanwhile, Washington's camp had a similar complaint that their backers didn't bother to vote because Harold was a shoo-in.

As a professional poll taker, I think last minute polls are useless. They are, moreover, extremely destructive, not only for the candidates but for the news media that employ them. There are few ways the average person can check the accuracy of a television station's news coverage of a particular story. But when a TV station runs a last minute poll, nothing is easier than to check the claims of the poll against the actual and nonsubjective results of the election.

The most significant use of a poll is to find out what people are thinking, not how they will vote in the next five minutes.

Consider a photograph of a mountain side. All seems serene as time is stopped at l/50th of a second. But in real time, you can hear the roar and see the avalanche in full fury. Woe to they who stand in its path.

So it is with polling. When a station dates the data of a snap poll the day or so before an election, they have a photograph of the numbers. But they have not analyzed them in real time. They do not realize that the landslide is really moving in another and more awesome direction.

An example of this was the first Adlai Stevenson-Jim Thompson battle for governor. Nearly all the last minute polls had statistics that showed Thompson winning by a mile. Thompson himself believed those figures and stopped campaigning for himself to help fellow Republicans on his slate.

My figures were similar, but I could see a tremendous potential for a huge turnout of Chicago Democrats, therefore I called the election too close to predict a winner. Of course, the result was a dead heat, with Thompson winning not by a mile, but by a millimeter.

April 1987/Illinois Issues/20

Another case to illustrate the same point was a congressional election in Indiana, in which an underdog was given almost no chance. Although I had similar numbers, the candidate was taking the same side on all the issues as the voters were talking, Because of this latent potential, I called it 50-50, and so it turned out. The underdog did finally lose, but by only 42 votes out all votes case in the entire whole congressional district. I mention these cases not to bolster my ego but to prove a point. People vote according to what they perceive in their houseeholds, not on what they see on a TV screen. Poll takers cannot move people to vote the way they want. They can, however, find out what it is that the voters want and advise their clients accordingly. A wise office seeker then has the option of either leading to press for action on a issue critical to voters or of presenting cogent reasons to oppose action. In the latter case, the reasons must win the respect — and the vote — of the constituency. It is still possible in America for a candidate to win elective office through respect.

There is an apt analogy between election races and horse races. A good jockey looks for open paths for his horse; he maneuvers to find hard when others are galloping through mud, and he tries to keep close to the rail to save ground. Similarly, a candidate uses a poll to find room to maneuver, to find good ground and to save his energies for what counts.

Getting back to Marshall McLuhan, is the medium the message? Not in the case of last minute polls.

The results of these polls have been dismal. This is probably not the fault of the polling concept, but of a misunderstanding of the dynamics of polling itself. Perhaps it is compounded by the penny-pinching budgeteers and a misconceived hope of winning viewers through honest, but understood figures.

Of course, there is no way to put restraints on last minute polling and its use by TV stations because polling and the revelations of polls (no matter how inacurate) are included under First Amendment rights, and indeed they should not be restrained.

But then, the system just doesn't work. Now what.

This failure gives us the clue we need for a resolution. Eventually, stations that continue to use the last minute poll, like the little boy who cried wolf, will be ignored because viewers will become tired of their inaccuracies and switch channels. That is the worst fate that can befall a televison station, and a just and proper solution.

Michael McKeon is head of McKeon & Associates, a national polling organization.

April 1987/Illinois Issues/21