Dewey Wins in Landslide! (I'm looking for cases of disastrously wrong political polls)

[Jackpine Radical]

And my point is that I want to have a look at the circumstances that made those polls so wrong.

Gimme a hand here, folks.

[mysuzuki2]

1. Bush wins in 2000?

[Jackpine Radical]

2. Hah.

I wonder, though--did the polls actually show Bush ahead? The exit polls certainly went for Gore before they were "fixed."

[The_Counsel]

5. Bush/Gore 2000

I think most polls had Bush with a slight edge in the week leading up to the election. They were all within margin of error, though--usually 3-5%.

Wanna hear something funny? The conventional wisdom the day before Election Day was that Bush would win the national popular vote, yet lose the more important Electoral College vote. The Bush campaign actually had lawyers ready to argue the point under that scenario...

[Nye Bevan]

3. The 2004 exit polls certainly generated some controversy here.

The 2004 election brought new attention to the issue of exit polls.[49] Discrepancies existed between early exit poll information and the officially reported results. These discrepancies led some, including British Prime Minister Tony Blair, to prematurely conclude that Kerry won the election.[50] Expert opinion was divided concerning what inferences should be drawn from the cited discrepancies.[51][52]

Mitofsky International, the company responsible for exit polling for the National Election Pool and its member news organizations, released a report detailing the 2004 election's exit polling.[53] At issue were the early release of some poll information, issues regarding correcting exit poll data using actual voter totals, and differences between exit polls and official results.

The NEP report stated that "the size of the average exit poll error ... was higher in 2004 than in previous years for which we have data." and that exit polling estimates overstated Kerry's share of the vote in 26 states by more than one standard error and overestimated Bush's share in 4 states by more than one standard error.[53] It concluded that these discrepancies between the exit polls and the official results were "most likely due to Kerry voters participating in the exit polls at a higher rate than Bush voters". Polling expert John Zogby later called this explanation "preposterous".[54] The NEP report further stated that "Exit polls do not support the allegations of fraud due to rigging of voting equipment. Our analysis of the difference between the vote count and the exit poll at each polling location in our sample has found no systematic differences for precincts using touch screen and optical scan voting equipment."[53]

A study performed by the Caltech / MIT Voting Technology Project concluded that "there is no evidence, based on exit polls, that electronic voting machines were used to steal the 2004 election for President Bush."[55] This study was criticized for using data that had been corrected to match the official count, and thus "essentially analyzing rounding error".[56] On December 5, 2004 Charles Stewart III of MIT released a revised report which, he said, used pre-corrected data.[57] Two days later, however, Warren Mitofsky, who had overseen the exit polling, stated that the pre-corrected data were proprietary and would not be released.[58]

One paper (and a follow-up book) concluded that discrepancies in the exit polls were evidence that the election results were off,[59][60] though others alleged this paper was unscientific.[61][62]

Following the 2004 election, researchers looked at ways in which polling methodologies might be flawed[63] and explored ways to improve polling in the future.[64]

http://en.wikipedia.org/wiki/2004_United_States_election_voting_controversies

[SidDithers]

4. "Martha Coakley will win on Tuesday. Bank on it"...nt

Sid

[starroute]

6. I think there was one that predicted a Roosevelt loss in 1936

I believe it was a matter of a conservative magazine polling its own readers.

But I only know about this from my mother telling me about it when I was a kid. I'll see if I can find something online. I'll edit this post to add it if I do.


On edit: That was simple.

http://en.wikipedia.org/wiki/United_States_presidential_election,_1936#Campaign

The 1936 poll showed that the Republican governor of Kansas, Alf Landon, would likely be the overwhelming winner. This seemed possible to some, as the Republicans had fared well in Maine, where the congressional and gubernatorial elections were then held in September, as opposed to the rest of the nation, where these elections were held in November along with the presidential election, as they are today. This seemed especially likely in light of the conventional wisdom, "As Maine goes, so goes the nation", a truism coined because Maine was regarded as a "bellwether" state which usually supported the winning candidate's party. However, only 24% of those sureveyed answered, and the magazine's readership was mainly wealthy people, who were more likely to be pro-Landon.

That same year, George Gallup, an advertising executive who had begun a scientific poll, predicted that Roosevelt would win the election, based on a quota sample of 50,000 people. He also predicted that the Literary Digest would be wrong. His correct predictions made public opinion polling a critical element of elections for journalists and indeed for politicians. The Gallup Poll would become a staple of future presidential elections and remains one of the most prominent election polling organizations to this day.