Well, not to be pedantic, but — putting aside bandwagonism — no one really loses because of the accuracy of a poll. If your question is: does anybody every lose an election because of race even though polls show them ahead, the answer is probably: not anymore. Though not that that article didn’t even mention Douglas Wilder.
You’re correct, I should have phrased it differently. A more accurate statement would have read something like “Bradley’s numbers weren’t impacted by the Bradley effect.”
Thought, it also depends a little bit on what you consider the Bradley effect to be. Is it that people are honest in the voting booth and dishonest to pollsters (the traditional definition), or is it that people think they’re being honest to pollsters, but when they get into the voting booth they re-evaluate the situation and, at the last minute, just can’t pull the lever for a minority candidate? If it’s the later, then Bradley could have lost due to the Bradley effect.
Also, it is possible to lose due to the accuracy of a poll if the poll leads voters (or even candidates) to behave in a way that alters vote totals. If Bradley’s non-racist supporters chose not to vote that day because the polls inaccurately predicted a landslide based on the racist voters, then the poll would have impacted the race.
I have a note on trying to poll for the Bradley effect that I’ll email you.
It seems like the difference between thinking you’re being honest to pollsters and not being honest to pollsters is pretty small. OK, it doesn’t sound small, but you’re essentially just lying to yourself in the first scenario.
Do you happen to know if polls have failed in situations where race is not a factor? Because nobody ever bothers to address the idea that the pollsters just screwed up.
October 20th, 2008 at 10:54 am
Well, not to be pedantic, but — putting aside bandwagonism — no one really loses because of the accuracy of a poll. If your question is: does anybody every lose an election because of race even though polls show them ahead, the answer is probably: not anymore. Though not that that article didn’t even mention Douglas Wilder.
[.]
October 20th, 2008 at 10:54 am
not->note
[.]
October 20th, 2008 at 11:56 am
You’re correct, I should have phrased it differently. A more accurate statement would have read something like “Bradley’s numbers weren’t impacted by the Bradley effect.”
Thought, it also depends a little bit on what you consider the Bradley effect to be. Is it that people are honest in the voting booth and dishonest to pollsters (the traditional definition), or is it that people think they’re being honest to pollsters, but when they get into the voting booth they re-evaluate the situation and, at the last minute, just can’t pull the lever for a minority candidate? If it’s the later, then Bradley could have lost due to the Bradley effect.
Also, it is possible to lose due to the accuracy of a poll if the poll leads voters (or even candidates) to behave in a way that alters vote totals. If Bradley’s non-racist supporters chose not to vote that day because the polls inaccurately predicted a landslide based on the racist voters, then the poll would have impacted the race.
I have a note on trying to poll for the Bradley effect that I’ll email you.
[.]
October 20th, 2008 at 2:20 pm
It seems like the difference between thinking you’re being honest to pollsters and not being honest to pollsters is pretty small. OK, it doesn’t sound small, but you’re essentially just lying to yourself in the first scenario.
Do you happen to know if polls have failed in situations where race is not a factor? Because nobody ever bothers to address the idea that the pollsters just screwed up.
[.]
October 20th, 2008 at 4:42 pm
I can inquire into situations where polls simply failed. People tend to think that the 2004 presidential polling failed.
[.]