Has it occurred to you that it was the opinion and exit polls that actually got it right, while it was the vote counts that were wrong? We know that more people INTENDED to vote for Al Gore in Florida in 2000, but were thwarted by all the shenanigans that went on. And all those touch screen voting machines that flipped or lost votes, as well as "starving" Democrat-leaning districts of voting machines in Ohio in 2004 may be what made the difference in Kerry vs. Bush. So, maybe the pollsters don't need to correct anything; the election officials do.

How come only in the last couple of cycles have their been an over-representation of Democrats? How come that did not happen 10 or more years ago? It seems like it is a completely invented phenomenon to explain why Bush won in 2004. Another explanation is fraud, mostly computer-centered.

It seems like OTM, which I see as generally fair and thorough, had a duty to at least discuss the study by Steven Freeman of Univ. of Penn, "The Unexplained Exit Poll Discrepancy" of 2004. This paper gives a much better analysis of recent exit poll failings than the pollsters themselves, needless to say.

Wow, wow, wow. Bob Garfield, whom I have heard give amazing, hard-hitting interviews, dropped the ball so hard, I think it fell through the floor.
How can you discuss the inaccurate exit polls of the last few elections and NOT at least bring up the possibility of election misconduct? There were machines in Ohio that register 10,000 votes in a row for Bush!
I expect much more from OTM, but now I"m wondering what other blind spots Bob might have.
I'm not saying you should say the election was stolen, but to present a one-sided argument about "What's wrong with exit polls" comes across as hopelessly naive.
Sloppy, sloppy reporting.
Exit polls are used in other countries to determine election fraud, not elections used to determine what is wrong with exit polls.
I hope this election we can actually have a real election, no Supreme Court, no voting machine "errors".

I have to agree with the above and concur that you blew it. No surprise though, since - as I've noted in my blog, NPR Check, NPR news has never seriously covered the clearly stolen election of 2000 and the likely Republican fraud in the 2004 election.

I am expecting huge amounts of voter fraud and irregularities this election, no matter who "wins." There are serious doubts associated with the outcomes of the last 4 presidential elections in particular.

Add my comment to say shame on you for not reporting what is well documented across the Web. There hasn't been an honest election in the US in the past ten years, and it has gotten progressively worse.

Even when there were landslides in the 2006 elections when the Congressman was exposed as a child molester in the last two weeks of the campaign he still almost "won". In a lot of other races Republicans voted in multiples of the total voter registration and Democrats showed no votes at all even when they endured great hardship to do so.

A partial list of these is to be found at http://www.freeforall.tv/ and that was just talking about Ohio.

absolutely infuriating that an organization as seemingly astute and plugged-in as yours has chosen the course of willful amnesia regarding the last two presidential elections. and I wonder why Joe Lenski, instead of hawing around with the 'sorry, we blew it' line of defense, doesn't just state flat out what he knows to be true. Joe's organization got it right both times, '00 and '04. Exit polling, proper done, is the most consistently accurate polling there is. There is a long, long history confirming that. Bush lost both elections. This collective failure to face the truth is not good for our nation, nor the world. you ruined half my day saturday by delivering that BS to my ears.

Good jorb Bob Garfield in repeating the classic MEDIA flaw of 2000.

The exit polls were right as the full recount showed.

The media blew it, and you just blew it again.

What an idiot.

I second Alan Woods's comment. The Florida polling would only be a valid example if Bush had actually won Florida. There is no question that Gore won Florida, and it was only a coup by the Supreme Court that prevented the true winner from taking office.

the discussion of the 'faulty' exit polls in the 2000 election ignores the fact that a count of the Florida ballots, by a consortium of news organizations several months after the election, proved that the polls were correct: Gore won the vote. A recent documentary by Dorothy Fadiman, "Stealing America: Vote By Vote," documents what took place in Ohio in 2004

The flaws of polling are less about age-skewing or party-skewing and more about the growing number of adults without landline phones. This is the flip side of the poll just before the Great Depression of 1929 that relied on landline phones that skewed to the rich; the poll result was that no one expected financial trouble soon! The election result didn't match the exit polls because the 2000 results were first skewed by Bush's cousin at Fox News then by the Supreme Court's Bush v. Gore decision not to count the vote, then by the distraction of 9/11 overriding NEWSWEEK and NEW YORK TIMES stories that Florida voted for Gore but the government failed to count it. The 2004 poll failed based on the facts in the Greg Palast book held by the victim of the "don't tase me, bro" incident in the same room as U.S. Senator John Kerry; Ohio's votes were never properly counted, especially provisional and absentee votes. Audit-resistant voting, whether by touchscreen or by optical scan, just separates voters from their votes even more.