US 2012 Presidential Election Polls

matthew8 · 17-10-2012 8:31pm

Not since Hayes beat Tilden in 1876 while losing by 3% of the popular vote has a candidate won by 1% of the popular vote lost the electoral college. That is why we should have faith in national polls. If you win the PV by 1%, you win the election. State polls are all over the place anyway and it's harder to get an idea of the momentum.

Ozymandius2011 · 18-10-2012 5:33am

Amerika wrote: »

17, October – Gallup - Likely Voters Trial Heat

Romney – 51%
Obama – 45%

Still too early for any happy thoughts from my side, as a lot can still happen in 3 weeks.

http://www.gallup.com/poll/157817/election-2012-likely-voters-trial-heat-obama-romney.aspx

(And Paddy Power just put Romney's odds at 2/1... OH MY!)

Health warning: it's a pre-debate poll and the RCP average only has Romney +0.4%. The IBD/TIPP polls have Obama ahead. Also the latest Rasmussen aggregate swing-state polls have Obama leading 50-47. Rasmussen have jut released polls in Nevada with Obama leading 50-47, ahead 1% in NH and most polls still show him ahead marginally in Wisconsin.

If Romney has a popular vote lead he may be getting it in the wrong places, such as in already deep-red Texas or unwinnable Massachusetts. If things don't change much I'm leaning towards a 2000 style outcome of Obama winning the E.C. and Romney the popular vote.

Ozymandius2011 · 18-10-2012 3:02pm

Rasmussen Ohio poll just out: Obama leads 49-48. Was taken after debate.

Amerika · 18-10-2012 3:14pm

I tend to believe Rasmussen over most the other pollsters. It’s all going to come down to Ohio, Florida, Iowa, Nevada, Wisconsin, New Hampshire, Virginia, North Carolina and Colorado. In Ohio, more importantly, Obama’s numbers have been dropping and Romney’s have been rising since the debates. From all indications that I’ve see, I think that polling trend will continue over the next two and a half weeks.

nagilum2 · 18-10-2012 5:40pm

Amerika wrote: »

I tend to believe Rasmussen over most the other pollsters. It’s all going to come down to Ohio, Florida, Iowa, Nevada, Wisconsin, New Hampshire, Virginia, North Carolina and Colorado. In Ohio, more importantly, Obama’s numbers have been dropping and Romney’s have been rising since the debates. From all indications that I’ve see, I think that polling trend will continue over the next two and a half weeks.

I think it's all about Ohio. If Obama hold Ohio, he'll almost certainly take the election, regardless of popular vote. It's going to be very very tough for the candidate that loses Ohio to get to 270.

Ozymandius2011 · 18-10-2012 5:50pm

The SurveyUSA poll that has Obama +3 in Ohio says that 18% of its sample said they have already voted and they have gone to Obama 57-38, while those not yet having voted were evenly split 43-43. SurveyUSA - unlike Rasmussen - also poll cellphone users who are seen as more likely to vote Democrat (being low income).

Duck Soup · 18-10-2012 8:41pm

Ozymandius2011 wrote: »

The SurveyUSA poll that has Obama +3 in Ohio says that 18% of its sample said they have already voted and they have gone to Obama 57-38, while those not yet having voted were evenly split 43-43. SurveyUSA - unlike Rasmussen - also poll cellphone users who are seen as more likely to vote Democrat (being low income).

Yep, early voting is almost worth its own thread, if it wasn't for the fact that early returns information is relatively difficult to come by, even though 40 states have already started voting. I don't think the states themselves are allowed to release the information so they're reliant on phone polling to ask who's voted and for whom.

Ohio Democrats are definitely voting in large numbers early, despite the attempts to restrict early voting days - including Souls-To-The-Polls final Sunday before the election - and also the billboards going up around Ohio's minority-strong areas about voter fraud leading to fines and imprisonment.

It could be that the voter suppression effort has actually backfired by riling up the minority voters in Ohio, who know they're being targetted and are animated to vote in a way that they wouldn't otherwise have been.

Ozymandius2011 · 19-10-2012 3:45am

NBC/Marist/WSJ polls today: Iowa Obama leads 8%, Wisconsin 6%. Iowa is a state where around 14% have already voted as of a few days ago.

On the other hand Romney leads 7% (!) on Gallup though all the other polls have it much closer (even Rasmussen). I think it's increasingly clear that Obama i much stronger in the swing states than the national popular vote and as such, a 2000-type Electoral College-popular vote split is becoming more likely.

Duck Soup · 19-10-2012 8:28am

Nate Silver over on FiveThirtyEight has Obama chances of winning up to 70.4% from 61.1% last Friday.

http://fivethirtyeight.blogs.nytimes.com/

It looks like the effects of the Veep and 2nd Presidential debates are starting to kick in.

hooradiation · 19-10-2012 11:46am

Duck Soup wrote: »

Nate Silver over on FiveThirtyEight has Obama chances of winning up to 70.4% from 61.1% last Friday.

http://fivethirtyeight.blogs.nytimes.com/

It looks like the effects of the Veep and 2nd Presidential debates are starting to kick in.

Also, Nate to Gallup - why you so crazy?

Goes a ways towards explaining why Gallup seems to be the outlier almost constantly when it comes to polling.

the tl;dr version for those that baulk at so much text.

To be clear, I would not recommend that you literally just disregard the Gallup poll. You should consider it — but consider it in context.

The context is that its most recent results differ substantially from the dozens of other state and national polls about the campaign. It’s much more likely that Gallup is wrong and everyone else is right than the other way around.

Ozymandius2011 · 19-10-2012 3:19pm

Following new Rasmussen polling from Virginia which has Romney leading 50-47, the RCP no-toss ups map is now just 281-257 for Obama. If it holds Obama will be the first Democrat to win election as President without a single Southern i.e. former Confederate, state.

Disasterous poll from Pennsylvania by Susquehanna has Romney leading 49-45 there. If Obama loses PA then he's definitely a gonner. It hasn't voted GOP for President since 1988. Apparently Romney's gains are largely in socially-conservative Western PA and the Philadelphia suburbs.

Duck Soup · 19-10-2012 3:21pm

The Thursday 18th October Reuters/Ipsos daily tracking poll has:

Obama 47%
Romney 44%

Take that, Gallup tracking poll!

The weighting for various polls enables people to pick the ones they like - which is why I prefer an aggregator such as Nate Silver - but as long as the weighting is constant, any and all polls should be able to show you trends.

Here's the Reuters/Ipsos Daily Tracking trend line.

Thursday, October 11, 2012: Obama 44% - Romney 47% (Romney up 3)

Friday, October 12, 2012: Obama 45% - Romney 46% [First after the Veep Debate] (Romney up 1)

Sunday, October 14, 2012: Obama 46% - Romney 45% (Obama up 1)

Monday, October 15, 2012: Obama 47% - Romney 45% (Obama up 2)

Tuesday, October 16, 2012: Obama 46% - Romney 43% (Obama up 3)

Wednesday, October 17, 2012: Obama 47% - Romney 44% [First after the 2nd Presidential Debate] (Obama up 3)

Thursday, October 18, 2012: Obama 47% - Romney 44% (Obama up 3)

So, if Reuters/Ipsos is to be believed, despite all the volatility in the polls, there's a pretty obvious trend of Romney peaking about the time of the Veep debate and the lead returning to Obama and holding constant - probably reinforced since the second debate - ever since.

Even if you don't subscribe to the absolute numbers, the underlying trend is still there. A high water point for Romney about 11th October and a modest drift back to Obama ever since, which in turn has settled and remained.

Duck Soup · 19-10-2012 3:32pm

Ozymandius2011 wrote: »

Following new Rasmussen polling from Virginia which has Romney leading 50-47, the RCP no-toss ups map is now just 281-257 for Obama. If it holds Obama will be the first Democrat to win election as President without a single Southern i.e. former Confederate, state.

Disasterous poll from Pennsylvania by Susquehanna has Romney leading 49-45 there. If Obama loses PA then he's definitely a gonner. It hasn't voted GOP for President since 1988. Apparently Romney's gains are largely in socially-conservative Western PA and the Philadelphia suburbs.

Against that, the new NBC/WSJ/Marist poll has Obama up 51-43 in Iowa and up 51-45 in Wisconsin.

http://firstread.nbcnews.com/_news/2012/10/18/14542135-polls-obama-holds-his-lead-in-iowa-wisconsin

I'd also point out that the Susquehanna poll was paid for by the Pennsylvania Republican Party.

http://www.politicspa.com/updated-senate-pagop-poll-romney-49-obama-45/42817/

Ozymandius2011 · 20-10-2012 4:24am

Wisconsin Rasmussen poll is 50:48 in favour of Obama. Romney leads 54-43 among early voters. Among those who have not yet voted Obama leads 50-49.

Black Swan · 20-10-2012 4:36am

The RCP average (7 to 18 October) very slightly in favour of Obama at this moment in time.

Obama = 47.1
Romney = 47.0
Spread = +0.1 Obama

For all practical purposes, the RCP poll average is a tie. Not sure what this means, given that it's been reported only 1-in-10 contacted reply to voter surveys, and the predominant method of contacting them is by automated messages to landlines, excluding most mobile-only households (estimated 26%), call screening, etc. Lots of room for systematic error and threats to external validity in measurement.

We may be set up for a "Dewey Wins" telephone poll failure, where the pollsters predict a win by either Obama or Romney, but just the opposite happens, and the news media relying on the polls look the fools.

Ozymandius2011 · 20-10-2012 7:50am

A rare lead (these days) for Obama in a Florida poll 47-46. Cuban-Americans break 48-40 for Romney, with other Hispanics breaking 67-27 for Romney.

Early voting update. The last time I posted on this the NC turnout by party-registration was about 2-1 GOP. Now it's 47-34 Dem. However the early ballot requests are close at 41-39 Dem-GOP. Blacks are now 27% of the turnout so far compared to 7% a week or two ago. However 53% of the turnout is ages 60+ and in the South many Democrats vote GOP in Presidential election and NC has always had a registered Dem plurality over the GOP. 68.8% of the turnout is white, 27% Black and 4% "Other".

In Iowa where 19.5% have already voted, the Dem-registered votes are leading the GOP 45-30 and the ballot requests are about the same. Unfortunately, Ohio doesn't have party-registration. In Florida its 45-39 in favour of GOP but about 40-40 amongst ballot requests.

jank · 20-10-2012 3:09pm

Brian? wrote: »

Yes, but I would like it to be more democratic.

By having mob rule? Because that is what a true democracy would be like.

Mjollnir · 20-10-2012 3:11pm

jank wrote: »

By having mob rule? Because that is what a true democracy would be like.

** facepalm **

Yes, because having mob rule was exactly what he said.

jank · 20-10-2012 3:13pm

I think it is all about Ohio at this stage.

Brian? · 20-10-2012 3:22pm

jank wrote: »

By having mob rule? Because that is what a true democracy would be like.

So eliminating the electoral college = mob rule?

Have some sense eh?

nagilum2 · 20-10-2012 3:30pm

jank wrote: »

By having mob rule? Because that is what a true democracy would be like.

Mjollnir wrote:

Yes, because having mob rule was exactly what he said.

Brian? wrote:

So eliminating the electoral college = mob rule?

Have some sense eh?

Not exactly mob rule, but jank wasn't totally off base.

The Electoral College is all about preventing regional dominance by a particular candidate from carrying them to the white house. The founders preferred a situation where a candidate won by capturing the most population weighted states, not just the largest population overall. In an exaggerated example, this would prevent a candidate from taking the presidency by winning 80-20 in their home region (in a sense, a regionally localized "mob") while losing 55-45 in all the others.

We may get a snapshot of that in this election, with Romney winning by mid-double-digits in the southern states, resulting in an overall popular vote majority, while Obama eeks out 3-4 point wins in the other regions, resulting in the EC win.

Ozymandius2011 · 20-10-2012 3:51pm

Regarding the Gallup poll: it now emerges they are using the 2010 turnout model, which assumes a White vote of 79% of the electorate. I personally think that is a mistake on their part. Furthermore, in 2008 Gallup overestimated Obama's margin of victory twofold, and the GOP's 2010 victory likewise twofold.

Gravis poll just out in Ohio: 47:47.

Brian? · 20-10-2012 4:56pm

nagilum2 wrote: »

Not exactly mob rule, but jank wasn't totally off base.

You're rationalising what he said. I'd appreciate him actually explaining himself.

The Electoral College is all about preventing regional dominance by a particular candidate from carrying them to the white house. The founders preferred a situation where a candidate won by capturing the most population weighted states, not just the largest population overall. In an exaggerated example, this would prevent a candidate from taking the presidency by winning 80-20 in their home region (in a sense, a regionally localized "mob") while losing 55-45 in all the others.

We may get a snapshot of that in this election, with Romney winning by mid-double-digits in the southern states, resulting in an overall popular vote majority, while Obama eeks out 3-4 point wins in the other regions, resulting in the EC win.

It's obvious to me though that the system skews the result of the popular vote by lowering turnout. Many dems won't bother voting in deep red states and vice versa with reps in blue states.

Frankly the EC is undemocratic because votes are weighted.

The senate elections are even worse. The electoral system is deeply flawed.

nagilum2 · 20-10-2012 6:21pm

Brian? wrote: »

You're rationalising what he said. I'd appreciate him actually explaining himself.

It's obvious to me though that the system skews the result of the popular vote by lowering turnout. Many dems won't bother voting in deep red states and vice versa with reps in blue states.

Frankly the EC is undemocratic because votes are weighted.

The senate elections are even worse. The electoral system is deeply flawed.

Well, we can agree to disagree then

I think the EC is a brilliant system that helps reinforce the notion that the federal US government is a government composed of all the member states. And it does what it is designed to do - prevent a candidate who performs extremely strong in one region from winning over a candidate that is preferred in a more widespread, but less pronounced fashion. The small states were given additional power to prevent politicians from only focusing on issues which affect the larger states. The fear was that without this power, politicians would completely ignore small states and only focus on big population centers. The EC ensures that a candidate can't just dominate a handful of those highly populated states and win. Yet, there are still enough electoral votes in the most populous 11 states that if they all go for one candidate, one could win the presidency on just those states over the votes of the other 39. In modern practice, it ensures that candidates must win those states that tend to have the most heterogeneous cross section of political beliefs, which I actually think is rather ideal.

I'm curious whether based on your argument, you feel Ireland should have fewer members of the European parliament? In fact the ratio for Germany is 1 MEP to every ~860k citizens vs Malta's 1 seat per ~67k citizens. Compare this to California's 1 EC vote per ~682k citizens vs Wyomings 1 EC vote per every ~192k citizens. The regressive proportionality for is actually much more pronounced in the EU parliament than in the US.

Manic Moran · 21-10-2012 1:09am

Brian? wrote: »

It's obvious to me though that the system skews the result of the popular vote by lowering turnout. Many dems won't bother voting in deep red states and vice versa with reps in blue states.

Frankly the EC is undemocratic because votes are weighted.

Welcome to a federation. One need not go far to find a system which give weight to individual states of the federation higher than that which simple populations would indicate. Germany, for example. (Bremen gets one vote in the Bundesrat per 200,000 people, North Rhine-Westphalia gets one vote per 3,000,000. (And even more so, all representatives from a German state must vote the same way in the Bundesrat, there is no such restriction in the US system). Like the US Congress, the Bundestag is split up on a seat-per-population basis.

The US is weighted more than most to the system being a federation of individual, self-governing, equal States. Although externally people may view the US as one nation, internally, it is far closer in concept to the EU, with the majority of decisions which affects the day-to-day lives of individuals being taken at the State level.

Not sure how it skews the vote by lowering the turnout. Unless a lower turnout in itself skews the vote (are people who are less likely to vote more likely to vote a certain way?)

The senate elections are even worse. The electoral system is deeply flawed.

What's wrong with the Senate elections? Every State gets two. Can't get any simpler than that, really. Even an American can understand the concept.

And it does what it is designed to do - prevent a candidate who performs extremely strong in one region from winning over a candidate that is preferred in a more widespread, but less pronounced fashion. The small states were given additional power to prevent politicians from only focusing on issues which affect the larger states. The fear was that without this power, politicians would completely ignore small states and only focus on big population centers. The EC ensures that a candidate can't just dominate a handful of those highly populated states and win

Fully agreed.

southsiderosie · 21-10-2012 2:12am

DonnaMarieAva wrote: »

OBAMA 2012 !!!!

K3lso wrote: »

GARY JOHNSON 2012 !!!

MOD NOTE:

This is a discussion board, not a sloganeering board. Please dial it down.

To everyone else: I'm a little worried that the EC discussion is taking away from the polling issue (although I agree that they are related), so I'd like to ask everyone to keep discussion of the EC tied to polling. If folks want to discuss the institutional merits of the EC, please feel free to start a new thread.

jank · 21-10-2012 8:29am

Mjollnir wrote: »

** facepalm **

Yes, because having mob rule was exactly what he said.

That is what "more" democracy is.

A republic normally has a constitution which protects the rights of all first and foremost regardless of majority.

A pure democratic country could vote to enslave the minority, in a constitutional republic that cannot happen. I would suggest to all to actually read what the difference is between a republic and a democracy.

http://www.lexrex.com/enlightened/AmericanIdeal/aspects/demrep.html

Simple minded quips like "I want more democracy" means nothing.

jank · 21-10-2012 8:34am

Brian? wrote: »

You're rationalising what he said. I'd appreciate him actually explaining himself.

How about explaining what you mean by "more" democracy.
Would eliminating the EC would that guarantee more democracy and if so how?
If it doesn't guarantee it why are you advocating it other than populism?

An example, do you think the EU should be run on the basis of a popular vote, after all wouldn't that be more democratic? :rolleyes:

Mjollnir · 21-10-2012 3:53pm

jank wrote: »

That is what "more" democracy is.

No, not until one can actually quantify what 'more' is and to what degree.

A republic normally has a constitution which protects the rights of all first and foremost regardless of majority.

A pure democratic country could vote to enslave the minority, in a constitutional republic that cannot happen. I would suggest to all to actually read what the difference is between a republic and a democracy.

http://www.lexrex.com/enlightened/AmericanIdeal/aspects/demrep.html

Simple minded quips like "I want more democracy" means nothing.[/QUOTE]

A. Spare us the remedial civics lessons, please. The US is a democratic republic. No has has even remotely indicated that it's a 'pure democracy' or that they want it that way.

B. What's simple minded, and transparently so, is the attempt to equate the statement, "Yes, but I would like it to be more democratic." with "I WANT MOB RULE!!!!!!!" and actually citing a statement that wasn't made. Rather intellectually dishonest, that.

No one's buying what you selling, so I would suggest that you kindly dispense with that nonsense.

Min · 21-10-2012 5:53pm

NBC/WSJ poll has Romney and Obama tied at 47%.

I see tweets about a poll putting Romney 4% ahead in Pennsylvania, where the Romney team have restarted their canvassing with Paul Ryan yesterday.

Polls would suggest, Florida, North Carolina and Virginia will go Romney, polls in the other batleground state are getting tighter.
New Hampshire moving into the Romney, the momentum seems to be still with Romney.

US 2012 Presidential Election Polls

Comments