In the first place, exit polls have not been historically thought to be unreliable. They usually can be counted on to pretty accurately reflect whom the voter has selected. They did receive a bum rap in Florida in Election 2000 because there was such a high level of spoiled ballots that year. In other words, the polls were correct in showing the numbers for the candidates that the voters thought they had voted for but they could not take into account that poorly designed ballots had caused the voters to incorrectly mark their actual ballots which of course changed the outcome of the vote.
Actually, exit polls have long been used as evidence by international organizations (including the US) to challenge rigged elections in foreign countries.
Now in both Florida and Ohio, we have exit polls that differ wildly from the actual vote. However, this time, we cannot go back and review the ballots to explain the discrepancy. We cannot do this because these two states relied heavily on DIEBOLD voting machine that do not have a paper trail to audit.
We knew coming into Election 2004 that electronic voting increased the possibility to rig elections:
(From http://www.crisispapers.org/essays/fixed-election.htm )The Republicans are well aware that exit polls can be used to warn us when "mischief is occurring:
"For example, in the highly suspicious Georgia senatorial and gubernatorial election of 2002, conducted entirely with touch-screen voting machines, the discrepancy between the pre-election polls and the election returns were beyond belief. The Democratic governor, Roy Barnes, who led Republican Sonny Perdue in the polls by 9 to 11 points, lost to Perdue by five points – a swing of about sixteen points. Max Cleland, ahead by five points,lost to Saxby Chamblis by six points.
These were shifts, respectively, of five and four margins of error. Talk to a statistician, and you will learn that the probabilities of such anomalies are vanishingly small. Is there, or was there, any way to validate the vote in those machines? Absolutely not. There is no independent record of the votes. That's just how the machines are designed."
Rather than blame exit polls -- or the bloggers who publish them -- and call for the discontinuation of these polls, we must insure that exit polls are widespread and fairly and impartially conducted so as to serve to remain as strong statistical evidence of fraud and manipulation. More importantly, we must insist on a paper trail for all elections. The Democratic Party let this issue slid by and blew the election this time around.
(From http://www.crisispapers.org/essays/fixed-election.htm )
"Soon after the polls closed in 2002, the exit polls from Voter News Service began to announce trends that were favorable to the Democrats. Then, suddenly, VNS told us that the results were “unreliable” and the entire election day operation was shut down.
How convenient for the GOP! And we have never received a clear explanation of what happened. Just the same message we had heard two years before in 2000: 'Get over it!'
In fact, in the past, exit polls have proven to be the most accurate polling methods. So they must be brought back as “checks” against the non-auditable machines. If the media's service, VNS, won't do it, then the left must demand an explanation of why it won't, and an alternative exit polling service must be put in place".
Will we continue to trust Republican-owned companies such as DIEBOLD to honor our most sacred right to freely elect our own officials?
The Republicans placed antigay marriage amendments on the ballot in eleven states this election. Call it Machiavellian or incredibly smart -- it did the trick to turn out their base.
We must start working now to place "Count Every Vote" (paper trails) amendments on every states ballots in the 2006 midterm election.
It will serve to not only ensure fair and free elections, but it will also turn out our base in an off year.