CO_Hoya
Golden Hoya (over 1000 posts)
Posts: 1,109
|
Post by CO_Hoya on Nov 7, 2006 21:58:16 GMT -5
Since I'm too lazy to do a Google search, thought I'd post my question here for all of the wonks: Do exit polls trend to Democrats? That is, do they lead to predictions that are incorrectly skewed toward Democrats? I seem to remember that being the case in 2004 (leading to early, erroneous predictions that Kerry would win), and I just saw a blurb on the NYTimes Caucus Blog saying that Fox News "will not be relying on exit poll data anymore to make its calls, but will rely on actual vote tabulations and historical models. They have found that the Democratic overrepresentation in the exit polls was 6 to 8 percent in many areas." And if this is so, why?
|
|
DFW HOYA
Platinum Hoya (over 5000 posts)
Posts: 5,777
|
Post by DFW HOYA on Nov 7, 2006 22:11:00 GMT -5
This was discussed on the Post election blog in the last hour. The feeling is that exit polls skew Democratic, perhaps because Republican voters are statistically less likely to participate.
The media pundits like to talk about this wave of change, but change is generally confined to the swing states. The Democrats still hold New York, for example, and very little was going to change that. Out here, the Republicans still hold Texas. The Democratic candidate for Texas governor is tracking at a whopping 29% of the vote. (Even Republicans in Maryland do better than that.)
|
|
CO_Hoya
Golden Hoya (over 1000 posts)
Posts: 1,109
|
Post by CO_Hoya on Nov 7, 2006 22:14:48 GMT -5
The feeling is that exit polls skew Democratic, perhaps because Republican voters are statistically less likely to participate. Again, and I'm not trying to be obtuse, why?
|
|
DFW HOYA
Platinum Hoya (over 5000 posts)
Posts: 5,777
|
Post by DFW HOYA on Nov 7, 2006 22:39:01 GMT -5
The feeling is that exit polls skew Democratic, perhaps because Republican voters are statistically less likely to participate. Again, and I'm not trying to be obtuse, why? Academics have been scratching their head for six years and can't figure it out, but it seems to be a real factor. I tend to think it's a sampling issue as to where and when exit polls are taken. And Joe Lieberman just declared victory in CT. How strange would it be if the independent Lieberman was the deciding factor in the balance of the Senate?
|
|
|
Post by jerseyhoya34 on Nov 7, 2006 22:44:09 GMT -5
He has said that he will caucus with the Democrats. Hopefully Republicans will not endorse flip-flopping on that issue as they seem to discourage it in any form.*
|
|
|
Post by StPetersburgHoya (Inactive) on Nov 7, 2006 23:48:35 GMT -5
I think that there are several issues with exit polling:
The big one is that it's not the sample size, its the sample location. You are collecting people at a time when they are extraordinarily unlikely to cooperate - sitting in a line, reading ballot initiatives, and wanting to get to/back to work or get home - they are going to be less likely to participate and also less likely to tell you the truth.
Also, in the past years I think the issue is that people are going to be more comfortable discussing issues that they feel strongly about it - votes for change or to protect something are more likely to skew towards those candidates or issues because people are going to be more willing to share those perspectives with exit pollers. What this means is that in elections that shape up with candidates that are running on platforms of change (i.e. 2004, 2006) are more likely to have supporters who share this information with exit pollers. This lead to seemingly weird results in 2004 where the early exit polls said that there were strong majorities for Kerry and bans on gay marriage.
|
|
|
Post by TrueHoyaBlue on Nov 8, 2006 10:18:18 GMT -5
While we probably won't know for certain until at least a few days from now, it appears that the exit polls were highly accurate in predicting this years actual results... well-placed adjustments? or just dumb luck?
|
|
SirSaxa
Silver Hoya (over 500 posts)
Posts: 747
|
Post by SirSaxa on Nov 8, 2006 11:59:28 GMT -5
I think that there are several issues with exit polling: The big one is that it's not the sample size, its the sample location. You are collecting people at a time when they are extraordinarily unlikely to cooperate - sitting in a line, reading ballot initiatives, and wanting to get to/back to work or get home - they are going to be less likely to participate and also less likely to tell you the truth. Actually, St.Pete, the exit polls are exactly what they say they are -- Polls taken when people are EXITING the polls. SO no one is "in line, reading ballot initiatives". It is illegal to contact people as they are in line waiting to vote. The reason that I heard about the errors in 2004 were largely related to the sample size techniques being inadequate to measure the extraordinarily higher turnout by the Republicans -- especially the extreme right wing christians. Exit polling has been fairly accurate over the years and has gotten better each year.
|
|
|
Post by StPetersburgHoya (Inactive) on Nov 8, 2006 13:44:57 GMT -5
I should have written "after" in that sentence. I've been asked questions after leaving a poll before - I hate people asking me what I just did. Its supposed to be a private matter and I'm not going to share it with a stranger.
|
|
DFW HOYA
Platinum Hoya (over 5000 posts)
Posts: 5,777
|
Post by DFW HOYA on Nov 8, 2006 13:52:20 GMT -5
The reason that I heard about the errors in 2004 were largely related to the sample size techniques being inadequate to measure the extraordinarily higher turnout by the Republicans -- especially the extreme right wing christians. The phrase "exteme right wing" or "extreme left wing" is hyperbole. There are comparatively very few on the true extemes of both parties--whether John Birch Society folks on the right or true socialists on the left. While the activists like to stir up the mailing lists by inferring their opponents are either a) plotting a theocracy or, b) Marxists in training, neither is the case. Evangelical Christians (Catholic or Protestant, regardless of denomination) vote consistently on social issues, less so on political ones. They're certainly more doctrinaire than extreme, but it's when people in one camp never get to know the other camp and deal with them on a regular basis that such fears get traction. Put another way, maybe there just needs to be a few more evangelicals in New York and a few more progressives in Florida for both sides to respect the other.
|
|
Boz
Blue & Gray (over 10,000 posts)
123 Fireballs!
Posts: 10,355
|
Post by Boz on Nov 8, 2006 14:38:47 GMT -5
He has said that he will caucus with the Democrats. Hopefully Republicans will not endorse flip-flopping on that issue as they seem to discourage it in any form.* True enough, but still, how great must it be to be Joe Lieberman today? How could you resist tugging at some chains and dangling the idea of going over to the other side.....at least for a little while as they're sorting out the mess in the Commonwealth? If I were Lieberman, I'd be walking down the hallways with a staffer right behind me holding up a sign with the words: "HOW YA' LIKE ME NOW, B**CHES!!!" I guess that's why he's a Senator, though, and I'm just this guy. ;D
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Nov 8, 2006 14:54:56 GMT -5
I guess that's why he's a Senator, though, and I'm just this guy. ;D With all due respect to Lt. Pete Mitchell and Lt. Tom Kozanski: "You can be my Senator anytime."
|
|
SirSaxa
Silver Hoya (over 500 posts)
Posts: 747
|
Post by SirSaxa on Nov 8, 2006 15:26:28 GMT -5
The phrase "exteme right wing" or "extreme left wing" is hyperbole. There are comparatively very few on the true extemes of both parties--whether John Birch Society folks on the right or true socialists on the left. While the activists like to stir up the mailing lists by inferring their opponents are either a) plotting a theocracy or, b) Marxists in training, neither is the case. Evangelical Christians (Catholic or Protestant, regardless of denomination) vote consistently on social issues, less so on political ones. They're certainly more doctrinaire than extreme, but it's when people in one camp never get to know the other camp and deal with them on a regular basis that such fears get traction. Put another way, maybe there just needs to be a few more evangelicals in New York and a few more progressives in Florida for both sides to respect the other. I am not sure how you separate voting on "social issues' and voting on "political ones". The problem that many people have with evangelicals is their virtual lock-step commitment to vote Republican no matter what. This time, however -- if the exit polls from yesterday are to believed -- something like 36% of evangelicals voted for Democrats indicating that they are finally beginning to realize that Republicans have been taking us all in the wrong direction. While the idea of everyone meeting and understanding one another is always good, the real problem that progressives have with evangelicals is they are trying to legislate behavior and they are actively trying to bash down the wall between church and state. Progressives believe that everyone has an absolute right to their own personal beliefs, but that churches -- and other religious institutions -- should not be getting involved with politics. As for understanding other parts of the country, maybe Texas is so isolated and insulated that they don't realize that Democrats didn't "hold NY", they actually elected a Democratic Governor for the first time in 12 years, and also sent more Democratic Representatives to Congress. Perhaps Texans are also unaware that NYC has had a Republican mayor for the last 14 years. I am not sure what Texans WOULD consider to be a "wave", but picking up 30+ seats in the House and almost certainly 6 seats in the Senate -- and control of both bodies -- as well as 6 more governorships, sure looks like a wave to most of the country. Even Texas' favorite son GWB referred to the election as a "thumping". And despite the Texas Republicans best (though legally questionable) efforts to gerrymander themselves into a "permanent majority", even Texans living in Tom DeLay's district had enough and voted to switch to a Democrat. DFW, I do appreciate your intention to convey the message that understanding one another is essential to the country making progress together -- and of course I agree. But when born-again Texans like Bush equate voting for Democrats to "the terrorists win", that is hardly a prescription for bi-partisanship or mutual understanding. It is not the progressives who have created the division in the country. And it was NYC that was attacked on 9/11. Those who have taken us to war in Iraq seem to forget that.
|
|
Nevada Hoya
Blue & Gray (over 10,000 posts)
Posts: 18,485
|
Post by Nevada Hoya on Nov 8, 2006 15:37:34 GMT -5
I still think the 2004 exit polls in Ohio were a conspiracy.
|
|