Tuesday, November 2, 2004
previous entry | main | next entry | TrackBack (2)
Open exit poll thread
I always favor more information over less information, so any exit poll info I get my hands on will be posted here.
However, please, please, PLEASE read Mark Blumenthal on the inherent uncertainty and limited utility of exit polls (particularly the early ones) before reading further. Hell, read what I wrote about this two years ago (and forgot about until James Joyner linked to it!!). Remember, when you're looking at exit polls, you're looking at raw sausage [Wonkette will love that analogy!!--ed.]
OK, done with that? Let the rumors, extrapolation, and mindless speculation commence!!
3:00 PM ET: Very strange -- Drudge had early figures from the National Election Pool posted. As I was looking at them, the screen refreshed, and poof, they were gone! Fortunately, Jonah Goldberg has posted them -- as has Wonkette.
Here's the full set of numbers that have been floating around (first number is Kerry, second is Bush):
The raw data has Kerry up by 20 points in Pennsylavania and up by 16 points in New Hampshire. That should tell you the size of the variance in these polls, because there's just no way Kerry wins by twenty points in Pennsylvania. Drudge says that the "early sample was based on a 59- 41 women to men ratio" -- which would partially explain those numbers. [59-41 for which states??!!--ed. Damned if I know -- though Cliff May has a silly theory for why this is true.]
UPDATE: Jonah has more:
Slate promises to post the numbers on their site, so be sure to check them out on a semi-regular basis.
3:25 PM: Now Wonkette has new numbers (first number is for Kerry):
Those numbers are all way too tight to extrapolate anything for anyone.
4:10 PM: See, this is why I'm glad danieldrezner.com's audience is so.... selective.
4:20 PM: Slate's first set of numbers -- which appear to be a mixture of morning and early afternoon polls:
4:40 PM: Wonkette has new numbers:
Drudge says, "One block from ground zero in NYC, 2 hour wait to vote..."
5:40: Slate now has the 4 PM exit polls [UPDATE: OK, these have now mysteriously disappeared from their web site -- may be due to the problem alluded to by Wonkette's source below]:
NRO's Kathryn Jean Lopez proffers the following set of numbers at 5:28 PM:
Both Drudge and NRO point out that early exit polls had Gore up in Florida by 3 and that didn't pan out as expected. This is true -- but if memory serves, those same polls had Bush winning the Electoral College pretty easily when you added up states -- Bush was winning in Iowa, Pennsylvania, and Wisconsin in the early exit polls of 2000.
6:05 PM: According to MSNBC, with "0% of precincts reporting," it's 61% to 38% for Bush nationwide!!! Seriously, I have no idea where those numbers are coming from.
UPDATE: Kudos to MSNBC for this page, which suggests that they'll be posting exit polls once the voting officially ends in each state.
6:06 PM: Scott Elliott says that, "My understanding is that exit polling does not include absentee and early voting. That is a very important point, given that as many as 20-30% of voters have already voted in some spots, and just re-emphasizes the worthlessness ofs exit polling." I don't think that's entirely correct -- I believe National Election Pool is trying to incorporate early voting, but they're doing it via phone polls -- less reliable than exit polls. Cleck here for more on early voting.
6:15 PM: Wonkette has fresh, hot, supple numbers:
Furthermore a source tells her, "There appear be problems with exits in the following states that could be tipping numbers toward kerry: MN, NH, VT, PA, VA, CT, DE. described only as 'serious' issues we're looking at. so i would not put too much faith in those results." UPDATE: Go check out Noam Scheiber on possible biases in exit polling and what they mean.
6:32 PM: Drudge now has Ohio tied, Kerry up by 2 in Florida and Minnesota, and up by 4 points in Wisconsin. I can't tell what he's saying about Pennsylvania, and Bush is up by seven in New Hampshire.
FINAL UPDATE: OK, go to this page at CNN or this one at MSNBC for
I feel confident (TradeSports) Kerry is going to win. I think there a lot of strong statistical trends his direction. There is no way the PA exits are gonna stick, but it would be great if it was a blowout.posted by: Jor on 11.02.04 at 03:30 PM [permalink]
Iowa Electronic Markets lurch towars Kerry. Started the day at 48.5%, dropped to about 47, now running past 55. I'd take all these early indications with a boulder of salt.posted by: Asclepius on 11.02.04 at 03:30 PM [permalink]
Slate and Wonkette sem about on a par for credibility...almost as low as the NYT
Seems to me that a pretty good way to protect the actual exit poll numbers would be to release a flood of fake leaks and leave non-subscribers wondering what is real and what isnt.posted by: GrantC on 11.02.04 at 03:30 PM [permalink]
More market stuff, Bush is TANKING on Tradesports. Exits + Turnout #'s (turnout being way more important) for Bush look really, really bad. That plus desperation tactics must be taking their toll.posted by: Jor on 11.02.04 at 03:30 PM [permalink]
It instills in me a lot of confidence when the exit polls spell it "Colarado".posted by: JJ on 11.02.04 at 03:30 PM [permalink]
Bush is tanking in the Iowa Electronic markets, now down in the high 30's. Kerry in the mid 60's. Either Kerry really is taking this or there are going to be some egg-faced "investors."posted by: Asclepius on 11.02.04 at 03:30 PM [permalink]
Oh god. It feels like the Game 7 of the ALCS for this Red Sox fan. Good god almighty, we just might pull it off!!!!!!!!!!posted by: pleasepleaseplease on 11.02.04 at 03:30 PM [permalink]
3:00 PM numbers had Wisconsin listed twice; 4:40 numbers had Florida listed twice. And are the Minnesota numbers for the whole state or just Minneapolis metro?
FWIW: Touch-screen voting has been a personal crusade of Cathy Cox, Georgia's Secretary of State. You'll never hear Jimmy Carter mention Georgia when he complains about touch-screen because Cox is a Democrat and a likely candidate for Governor in two years. Anyway, I got through a line that stretched out the door and around an outdoor parking lot, easily 200+ people long, in about 40 minutes. I don't know enough about the tech to say whether fears that touch screen loses votes or is prone to system crashes are justified, but it is a lot faster and consequently popular here.posted by: Zathras on 11.02.04 at 03:30 PM [permalink]
TradeSports and the IEM remind me of October 1929 - pity the folks who cant get through the logjam .... Remember that scene in Trading Places where Ackroid and Murphy start the selling frenzy?posted by: TexasToast on 11.02.04 at 03:30 PM [permalink]
PS - I was one of three voters (yawn) at 10 AM in my pricinct - in Fort Worth.posted by: TexasToast on 11.02.04 at 03:30 PM [permalink]
My intown Atlanta precinct had a long wait this morning. All due to the lines to get the id check, and to check name against the voting rolls. People were breezing through the ballot, and there were polling screens standing empty.posted by: Appalled Moderate on 11.02.04 at 03:30 PM [permalink]
Slate and Wonkette seem about on a par for credibility...almost as low as the NYT
Isn’t Wonkette pretty much a sex scandal blogger? In which case why is Drezner quoting unsourced “exit poll” numbers from her?
posted by: Thorley Winston on 11.02.04 at 03:30 PM [permalink]
I think it's by now fairly clear to everyone that the exit polling was skewed toward Kerry beyond all reason. I regard this as wishful thinking by the poollers.
I think it also fairly well clear why Kerry did fairly well in the earlier returns; The first to return numbers is going to be the better-conneted urban voting districts, where Democrats are likley to do better anyway.posted by: Bithead on 11.02.04 at 03:30 PM [permalink]
Post a Comment: