The only poll that counts
After nearly four years of campaigning, nearly two dozen candidates, millions of dollars in advertising, and thousands of charges and countercharges, it all comes down to this: No one knows what will happen on Election Day.
Will Barack Obama maintain his five to 10 point lead in the polls and win a landslide victory on Tuesday? No one knows.
Will John McCain benefit from lingering doubts about Obama's background and pick up enough key states to win an Electoral College victory even if he loses the popular vote? No one knows.
Do the polls, which on Friday showed Obama leading by anywhere from 3 points to 11 points, have any predictive value at all? No one knows.
Is Montana leaning 4 points to Obama as a Montana State poll declared on Oct. 20, or 4 points to McCain as a Rasmussen Poll showed on Oct. 29? No one knows.
Is critical Pennsylvania within the margin of error as one poll showed on Oct. 28? Or a 14-point blowout for Obama as another poll showed the day before? Need we say it? No one knows.
Frankly, only a fool would predict the outcome of Tuesday's election - there are far too many unknowns, and far too much at stake for anyone to be confident.
Certainly, there is potential for a "Dewey Defeats Truman" style debacle for pollsters in this election, but that potential exists every four years, and almost always goes untapped. The circumstances that allowed Truman's come-from-behind victory to go unnoticed by pollsters don't exist anymore, but new circumstances do exist that could compromise the validity of the polls in other ways.
One problem is the widely reported Bradley effect - so named because Los Angeles Mayor Tom Bradley was far ahead in all polls leading up to election day, but lost by a considerable measure in the actual voting. Pollsters concluded that respondents had covered up their true intentions because they didn't want to reveal they were voting against the black candidate, Bradley.
It is not necessary, however, to conclude that people lied to the pollsters about their vote. Instead, you just have to assume that a substantial number of anti-Bradley voters refused to answer the pollsters' questions at all, leaving the sample of those who actually participated inevitably skewed to favor Bradley.
The same thing could very easily be taking place in the Obama-McCain race. Most Obama supporters are enthusiastically in favor of him, and proudly declare their support whenever possible. McCain's support is not necessarily so enthusiastic, and indeed many of his votes may really be anti-Obama votes. It is also possible that a lot of the anti-Obama votes are anti-establishment votes and anti-mainstream-media votes. Those people may just not wish to participate in polling which they see as biased in the first place, creating a self-fulfilling prophecy where the outcome is skewed toward the candidate they oppose.
Or Obama could be riding a tide of change (in the wake of the failed Bush presidency) that would have swept him to victory no matter what McCain did or said. No one knows.
So don't let your decision to vote or not vote be based on any expectation created by hearing about polling numbers on Election Day. Even exit polls aren't accurate, or else John Kerry would be running for re-election on Tuesday.
The only thing anyone knows for sure is that your vote matters, and so does this election.
. Frank Miele is managing editor of the Daily Inter Lake. E-mail responses may be sent to edit@dailyinterlake.com