“When I pull this thing off, I have a request for my opponent,” John McCain said at a rally [in New Mexico Saturday]. “I want him to save that manuscript of his inaugural address and donate it to the Smithsonian, so they can put it right next to the Chicago paper that says ‘Dewey defeats Truman.’”
Let's set aside for a moment the fact that that manuscript does not actually exist, but was merely a public literary device from a Democrat unaffiliated with Obama's campaign months ago. The Washington Post explains the Truman reference, and joins McCain in questioning whether or not the polls are accurate:
Could the polls be wrong? Sen. John McCain and his allies say that they are. The country, they say, could be headed to a 2008 version of the famous 1948 upset election, with McCain in the role of Harry S. Truman and Sen. Barack Obama as Thomas E. Dewey, lulled into overconfidence by inaccurate polls.
I would suggest that Senator McCain and the Post staff bone up a little on their recent American history. The 1948 polls that showed Dewey ahead of Truman were actually accurate. The problem was that pollsters stopped polling a week ahead of election day, and Truman's barnstorming whistle-stop tour around the country lambasting the "do nothing" Congress all week had a huge impact.
It's vital to remember: polls don't predict what will happen in an election several days or weeks down the line; they only tell you what would happen if the election were held at the time the poll was in the field. The Tribune headline got the story wrong in 1948 because it relied on out-of-date polls, not because those polls were wrong. A week prior to Election Day, Dewey WOULD have defeated Truman. But things change.
In 2008, out-of-date polls is the absolute last of our worries. Barring a severe backlash from tonight's infomercial or maaaybe Osama bin Laden's capture, Senator Obama will win this race and handily so; if not in a popular vote landslide, then certainly in an electoral bath.