AP-GfK poll shows White House race still tight (Obama 47% Romney 46%)

Associated Press ^ | 8/22/2012 | Josh Lederman and Tom Raum

[middlegeorgian]

For all the attention it got, Republican Mitt Romney's selection of Rep. Paul Ryan of Wisconsin as his running mate has not altered the race against President Barack Obama. The campaign remains neck and neck with less than three months to go, a new AP-GfK poll shows.

Overall, 47 percent of registered voters said they planned to back Obama and Vice President Joe Biden in November, while 46 percent favored Romney and Ryan. That's not much changed from a June AP-GfK survey, when the split was 47 percent for the president to 44 percent for Romney.

At the same time, there's a far wider gap when people were asked who they thought would win. Some 58 percent of adults said they expected Obama to be re-elected, while just 32 percent said they thought he'd be voted out of office.

[middlegeorgian]

Good news for Romney as AP-GFK poll is one of the more left leaning polling outfits. AP-GFK showed Democrats getting more support throughout 2010 than most pollsters. I couldn't find any D/R/I sample.

[Venturer]

It’s going to stay close too.

45% don’t pay taxes, Blacks ,queers,Hispanics , and women voting for the man who buys their vote by borrowing money, to keep them bought. They aren’t goging to vote to stop the freebies.

The Romney votes will come from those of us who are paying the bills. These people voting for the freebies do not realise we are reaching the tipping point.Or they don’t care.

[mazda77]

Just more shake-n-bake bullcrap. I have typed what I consider to be the real methodology so many times I am going to type a ready made explanation of these public polls paid for by someone else with an agenda so all can see just how easily this crap goes on for news and not real public opinion.

This is right off their site;

The AP-GfK Poll is based on a nationally representative RDD sample of at least 1,000 adults, ages 18 and older, living in the 50 states. The landline RDD sample of households called is stratified, or divided, by census region with targets set for the number of complete calls per region. Cell numbers are not geographically stratified, other
than in selecting the initial sample.

[cuban leaf]

What’s the difference between a Romney supporter and an Obama supporter?

Romney supporters sign their checks on the front.

[VanDeKoik]

It isnt going to be close.

But how else will a polling outfit make $$$$ if they dont keep the drama high?

They have to be right in late October when they have a reputation to maintain.

[WILLIALAL]

47% Dem

41% Rep

[TruthShallSetYouFree]

Despite the AP’s attempt to paint this as good news for Obama, the trend is our friend. First of all, this is a survey of registered voters, not likely voters. The former always skews in democrats’ favor, while the latter more accurately predicts real voting patterns. And, second of all, Obama has not managed to capture any more than the 47% he had two months ago, while Romney moved up from 44% to 46%. So Romney gained 2% in two months, while Obama was stagnant. Extending the trend line to election day would leave Obama mired at 47% while Romney passes him.

The total of 93% leaves 7% for Undecided or Other. After nearly 4 years, everybody knows what Obama is like. He’s not going to capture much of the remainder.

Of course, the overall percentage is merely a sentiment indicator, and not a predictor of who will win the Electoral College, the only vote that matters. Obama could (and probably will) win liberal enclaves like California, New York, and Massachusetts by very large margins. But it doesn’t matter how much he wins those states by. You get the same number of their Electoral College votes whether you win by one vote or one million. He could win big in those states and lose small in Florida, Ohio, Wisconsin, and other swing states, maybe even winning the popular vote while losing the election. God willing.

[profit_guy]

D31, R23, I30

D Strong 15, D Moderate 16

R Strong 9, R Moderate 14

http://ap-gfkpoll.com/main/wp-content/uploads/2012/08/AP-GfK-Poll-August-2012-Topline-FINAL_politics.pdf

typical lefty bias

[WILLIALAL]

When leaners are not forced to pick either Dem or Rep the totals are:

Dem: 31
Rep: 23
Ind: 30

[nhwingut]

So basically polled 9% real republicans and Obama leads by 1. LOL!

I’m guessing the exit polls will show more than 9% strong republicans will show up in November.

[PapaNew]

AP-GfK poll shows White House race still tight

There's either a major disconnect between dissatisfaction and voting or between the Gallup poll and the AP-GfK poll (or both).

http://nation.foxnews.com/2012-polls/2012/08/21/obama-slaughtered-latest-gallup-poll

[GnL]

Two words: registered voters.

[Buckeye McFrog]

Yep, sadly 45% seems to be the absolute floor for Socialists these days. And with the stunning economic ignorance of almost everyone under 30 I meet I suspect that will pass the 50% threshold within my lifetime.

[profit_guy]

Plus this poll is further doctored by using 68% Whites, when the actual % of Whites in the population is 72.4%

[SharpRightTurn]

Associated Propoganda continues to use registered voters rather than likely voters, thus intentionally factoring out voter enthusiasm and intensity (one of the reasons they didn’t see the 2010 results coming).

Of course, the Dims will have to get their “registered voters” to the polls during a smaller window than Republicans since so many Dims don’t get up before noon, then have to make a run to the convenience store for beer, cigs, and lottery tickets, and then a little afternoon partying. That’s always why the Dims try to keep the polls open after closing time (which also allows the dead voters to appear).

[Uncle Miltie]

Wow. That’s about a +7% under-sampling of Republicans. Plus, it’s “Registered” not “Likely” voters.

This is gonna be a blowout. I’ll predict 55% / 45% Romney / Obama.

[andy58-in-nh]

All of these polls are making an identical error: they are using 2008 turnout numbers as a baseline. It is far more likely that the 2010 election will be a better guide. Anyone who hopefully believes that only 23% of the electorate on November 6th will be Republican is in for a rude awakening.

[andy58-in-nh]

Polls rarely use "likely voters" before September of an election year. Their common mistake is using 2008 as a baseline for a registered voter sample; that year saw a drop in Republican registration and a rise in Democrat rolls, largely due to a lack of GOP enthusiasm and the organizing efforts of Obama's team, which brought in a large number of young first-time voters.

Since that time, registrations have reverted back to the earlier trend: a drop in Democrats, a recovery by Republicans and a growth in independents. The only major pollsters who appear to have incorporated this reality into their survey methodologies to any extent are Rasmussen (R) and Public Policy Polling (D), the latter of which still oversamples Democrats, but by a smaller margin than others.

[expat2]

Registered voters, too — not likely voters.

[Credo]

32-23 D to R. No bias in that sampling.