2016 Postmortem
Related: About this forumGallup claims Romney is winning early vote
http://www.slate.com/blogs/weigel/2012/10/30/this_early_vote_calculation_from_gallup_looks_terrible_for_obama.html?tid=sm_tw_button_toolbarSomeone is going to have major egg on their faces come election day, and I hope it's Gallup.
Response to smorkingapple (Original post)
Post removed
smorkingapple
(827 posts)RightBWrong
(26 posts)....is posting info from an actual article from a decent site like Slate trolling?
still_one
(91,807 posts)Case that is against du policy
I personally believe gallop has been screwed up for sometime now, and other polls contradict it also
phleshdef
(11,936 posts)smorkingapple
(827 posts)phleshdef
(11,936 posts)Gallup doesn't necessarily have a partisan skew. They just get it wrong, period.
still_one
(91,807 posts)Definitely wrong, especially compared to other polls
Lex
(34,108 posts)WI_DEM
(33,497 posts)winning early voters overwhelmingly?
FBaggins
(26,681 posts)For a few reasons:
* They tend to be based on small sub-samples with high error margins
* We have little track record at identifying whether early voting adds much to turnout (or instead just has some people voting a little early who would have voted anyway).
* What if one party focused on absentee voting while the other spent their GOTV efforts on early voting (where both is available of course)?
* Most importantly... I've watched this go on for too long to be fooled again. Take a look at all of the early voting threads here back in 2010. They were "proof" that the polls were wrong and this "enthusiasm gap" thing was nonsense. The election was going to go much better for us than people feared based on those phony polls...
... instead they were much worse.
The point is that both sides can find early data that they use to claim they're outperforming where it counts. I'm not buying it absent firmer evidence.
Coexist
(24,542 posts)octoberlib
(14,971 posts)berni_mccoy
(23,018 posts)Look at these totals:
Early Voters By Age:
18-29: 7%
30-49: 11%
50-64: 16%
65+: 26%
That totals 60%
What age are the other 40%?
Same is true by Region (only total 59%) and by Education (62%).
Numbers don't add up.
Furthermore, early voters by party id are 19% Repub; 12% Independent and 15% Democratic. That's 46% total, not to mention major oversampling of Repubs.
Poll is bogus.
FBaggins
(26,681 posts)It's not a breakdown of those who voted early (i.e., 7% of early voters were 18-29)... it was a report on what proportion of each demographic has already voted (i.e., 7% of 18-29 year olds have already voted).
berni_mccoy
(23,018 posts)and something they shouldn't be doing in an Early Voting poll. They are boosting Repub numbers by doing so as we all know Democratic request for early voting or absentee ballots have been beating Republican requests by 2-1 on average (where it is tracked). Gallup is adjusting the numbers by party id, something which they should not be doing in this poll.
FBaggins
(26,681 posts)It doesn't have anything to do with weighting.
It just tells you that different age groups are more/less likely to vote early than others.
as we all know Democratic request for early voting or absentee ballots have been beating Republican requests by 2-1 on average
Actually, we don't know that. There are some places where we're doing very well, and you'll see those on progressive sites cited as the standard for this year... then there are some places where we're badly underperforming 2008 and you'll only see those on RW sites (also claiming that they prove the election is in the bag).
We saw the same game played in 2010. Don't fall for it. The early numbers don't tell us much of anything... but it's all they have to report on now so that's how the game is played.
berni_mccoy
(23,018 posts)The only way you get the total to be 15% of all registered voters who have already voted with the breakdown by age or party idea that they have is if each segment is equally weighted.
That is, the average of percentages by age, 7, 11, 16, and 26 is 15. Likewise the average of percentages by party id, 19, 12 and 15 is 15 as well. That means, Gallup must have equally weighted each group.
It's highly unlikely these groups were equal in the sample.
FBaggins
(26,681 posts)They say that 13% of the electorate is 18-29 and that 7% of them have voted. 7% of 13% is .91%
They say that 34% of the electorate is 30-49 and that 11% of them have voted. 11% of 34% is 3.74%
They say that 31% of the electorate is 50-64 and that 16% of them have voted. 16% of 31% is 4.96%
They say that 22% of the electorate is 65 or older and that 26% of them have voted. 26% of 22% is 5.72 %
5.72% + 4.96% + 3.74% + 0.91% = 15.33%
msrizzo
(796 posts)Gallup is showing Romney will a 5 to 6 point lead over Obama so of course they voters they sampled will show Romney leading in the early voting....the majority of the people that pass their likely voter screen are Romney voters, right?
abumbyanyothername
(2,711 posts)if a person answers yes to the question, have you already voted?
skeewee08
(1,983 posts)President O is leading in early voting in Ohio, Iowa & Nevada....
obamanut2012
(25,869 posts)HockeyMom
(14,337 posts)Those areas are VERY heavy Democratic areas. People have posted that the lines have been out the doors. What you would have to conclude from that is that registered DEMOCRATS have switched to Romney.
Really? Wanna buy a bridge real cheap? I am in NY right now and cannot vote in early voting in Florida. I WILL vote for Obama on November 6th, because I didn't trust voting absentee ballot. JFK is supposed to open tomorrow and my flight isn't until Sunday. My ONLY desire to go back to Florida is to VOTE.
Don't they know that we Dems are very fired up and very aware we have to do everything we can to vote if we live in a swing state.
Alekei_Firebird
(320 posts)Is it any surprise that their EV poll shows the same thing?
Every single poll in the swing states except for CO shows a big Democratic advantage in EV. I don't know why we're fussing about an outlier poll.
theinquisitivechad
(322 posts)None of this squares up with the tremendous early voting leads we have seen for Obama in some swing states. Not surprising.
mzmolly
(50,952 posts)vs. ACTUAL early voters.
kstewart33
(6,551 posts)Gallup might be wrong, they might be right.
We can choose to ignore the post or deal with it. But automatically calling Sapple a troll for the post, ain't right.
FBaggins
(26,681 posts)It's a defense mechanism.
OilemFirchen
(7,143 posts)It's a reflection, if even accurate, of national polling trends, and, as the subhead indicates, "Early voting highest in the West and among seniors".
There's no breakdown by state, especially swing states, where Obama has a clear advantage in early voting.
Lex
(34,108 posts)where clearly the President's team has been concentrating his early voting ground game.
RomneyLies
(3,333 posts)Zogby is happy that Gallup has been so bad this cycle.
MSMITH33156
(879 posts)This is not just the swing states, this is all states. Here are the states WITHOUT early voting:
Alabama
Connecticut
Delaware
Kentucky
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
New Hampshire
New York
Oregon
Pennsylvania
Rhode Island
South Carolina
Virginia
Washington
That is 5 red states (Alabama, Kentucky, Mississippi, Missouri, South Carolina), totaling 47 Electoral Votes.
That is 10 blue states (Connecticut, Delaware, Massachusetts, Michigan, Minnesota, New York, Oregon, Pennsylania, Rhode Island, Washington), totaling 119 electoral votes.
That is 2 swing states (New Hampshire and Virginia), totaling 17 electoral votes.
Simply put, there has been a lot more early voting in red states, so if you look at the nation as a whole, you would expect Romney to be ahead in early voting. Almost half of Obama's strong states by electoral vote haven't voted yet.
CreekDog
(46,192 posts)many, many people have already voted in those states.
MSMITH33156
(879 posts)early voting totals though? You can even take those out, and the point still holds.
CreekDog
(46,192 posts)you can't compare Washington and Oregon (who have many, many voters have voted, are voting as we speak) with New York and Pennsylvania where you really can't vote until election day, in person.
you just have to take those states off your point and then it will be more accurate.
also, i think you're wrong to suggest that votes are tallied for candidates prior to the end of the election. don't think that's allowed anywhere.
MSMITH33156
(879 posts)in almost all the states, with few exceptions.
My question was actual more about the asking mechanism. How did Gallup ask the question? Did they ask it in a way that would encompass early and absentee voters, or did they ask it in a way that would make people that have absentee voted not count in their survey? That's what I meant. FWIW, here are all the early voting state information:
http://nass.org/index.php?option=com_content&view=article&id=223&Itemid=391
BlueStreak
(8,377 posts)augie-va
(7 posts)and I find this troubling...
"But... in 2008, Obama was winning this vote. An identical Gallup study taken around the same time gave Obama a 53-43 lead with early voters and a 50-44 lead on voters who would wait for election day."
-Slate
...and whether you choose to believe it's tampering, incompetence or spin bias, it is -along with the RAS poll- a critical part of the news cycle if you're listening past Sandy updates.
Nate called yesterday's polling for BO, mediocre; a word that will doubtless be employed again on Friday when the job report is released ( about the same as September and unemployment falling .1%)
fugop
(1,828 posts)This is exactly right. Gallup is being misleading (as usual), as is every pundit who announces this bullshit as a sign of where the election is going. Think of the population of the blue states that have no early voting. It's ludicrous. (Plus, as I understand it, this Gallup thing also included those who SAY they intend to early vote, even if they haven't yet?).
Ridiculous. Tuesday can't fucking get here fast enough.
Jennicut
(25,415 posts)That was 3 days ago.
http://www.reuters.com/article/2012/10/27/us-usa-campaign-voting-idUSBRE89P15920121027
"Obama leads Romney 54 percent to 39 percent among voters who already have cast ballots, according to Reuters/Ipsos polling data compiled in recent weeks. The sample size of early voters is 960 people with a credibility interval of plus or minus 3.5 percentage points."
Gallup compiled those who voted or "planned to vote". Big difference between using data only from those that actually did early vote. Gallup is crap.
Rosa Luxemburg
(28,627 posts)progressivebydesign
(19,458 posts)Gallup is such a right wing tool.
The people in line voting this week are the real Early Voters in our minds, not people on the phone that they polled who say they "plan to vote" before election day. Without a breakdown of who they asked, and the States, etc. it makes little sense. But the media is clinging to it desperately.
cash__whatiwant
(396 posts)Doctor Jack
(3,072 posts)Doesn't mean shit for Ohio, Florida, and North Carolina
SLCLiberal
(102 posts)Gallup includes those who "plan to vote early" in their poll and then count those as early voters?
Gotcha.