Articles and Analysis


US: Afghanistan, Wikileaks (USA Today/Gallup 7/27-8/1)

Topics: National , poll

USA Today / Gallup
7/27-8/1/10; 1,208 adults, 3% margin of error
Mode: Live telephone interviews
(Gallup release, USA Today release)
Update: Full Obama approval here.


Obama Job Approval
41% Approve, 53% Disapprove (chart)

Thinking now about U.S. military action in Afghanistan that began in October 2001, do you think the United States made a mistake in sending military forces to Afghanistan, or not?
43% Yes, a mistake, 52% No, not

In general, how would you say things are going for the U.S. in Afghanistan -
1% Very well, 33% Moderately well, 42% Moderately badly, 20% Very badly

If you had to choose, which do you think is better for the U.S. --
38% Keep troops in Afghanistan until situation gets better
57% Set time-table for removing troops from Afghanistan

As you may know, the documents about the war in Afghanistan were leaked to a website called "Wikileaks." Do you think it was right or wrong for that website to publish those documents?
25% Right, 66% Wrong



Its time to end all of these Republican Wars right now! Newt Gingrich told Fox News he wants Obama to attack Iran and N-Korea.When is the American people going to wake up,because this is all your going to get from Republicans,more Wars and more Debt to pay for it.



Why didn't Gallup ask how many respondents actually read the documents?


Field Marshal:

New low for Obama on this poll. 41% on adults.


What does having read all 66,000 documents have to do with the questions asked?



Youch! 41/53?




I wonder how much the lawsuit against AZ slammed Obama's approval down in this poll? This is incredibly low considering it is of all adults - not just registered or likely voters.

I also wonder how long his approval will stay this low. Everyone has been talking about how this election wouldn't be so bad since Obama's approval was hovering near 50% but if it hovers closer to 40% then it is a whole new ball game.

If we go into November with his approval at 40%, congressional approval at 12%, and unemployment at 10%, just how bad could it actually get for democrats?



There is no way Obama approval rating can be that low,because he would have to be getting only 28% of the White vote to have numbers that bad.



The only way to show a accurate poll of Obama, would be if these pollsters show the Demographics of they're polling,because if Obama is getting over 90% of the Black vote,and over 60% of the Latino vote,it means he would only need 38% of the White vote to reach 50% of the total vote.That's why this poll and other polls should show the Demographics, because that's the only way you can tell if these polls are accurate.




What are you using for your racial model that tells you that a 41% approval equates to an approval of 28% by whites? I'm interested to see the math you used for that.



If the polls stay where they are now, it'll be an absolute bloodbath, CompCon. Looking at yesterday's generic ballot thread, I can't believe how many poll deniers are out there.

RCP tabs all the same polls as Pollster.com with the exception of the three internet polls - Yougov, Zogby and that one new guy in the block - Politico poll.

With the exception of Gallup and a single, one point TimeMag poll, you have to go back eight weeks to find a poll that had Dems ahead on generic ballot.


Current total based on the latest six polls is GOP +6.

In 94 GOP won the popular House vote by 6% and they ended up with a 6% advantage in seats. One barely relevant point is that current House composition by party is almost identical to what it was in the 93-94 House.

But only one the six polls in RCP average are LV's. The rest are RV samples. Even in a high Dem enthusiasm/low GOP enthusiasm mid-term year, Repubs turn out at higher ratios than what RV polls indicate. RV samples underestimate Repubs by anywhere from 3 to 10%. If I go on the low end of that adjustment and only add 4.5%, the lead goes to 10.5.

In 2008 Dems won popular House vote by 10.65% but the seat count by 18%. The reason for divergence between popular vote and seat count in 08 is that once the lead gets big enough, the number of seats available to the losing party shrinks at accelerated rates. For example if one party has a 70/30 lead, the smaller party can only expect to win less than 5% rather than something close to 30%. That's because for the smaller party to win in a particular district, it would need to have a 40% inherent advantage in that particular district. And there are only so many districts where one party has 40% inherent advantage. It is similar to Electoral College Votes in a presidential election. If the popular vote is split 60/40 between two candidates, the one that receives 40% would only get maybe 10-15% of the EV's.

So based on CURRENT polling, and applying only a very lowball LV adjustment of 4.5%, 2010 looks like the mirror image of 2008. GOP ends up with about 260 seats and Dems with about 175.

If the election were held today, I have no doubt GOP would do at least that well. But even if the landscape stays exactly the same until Nov, I would still expect for the polls to tighten a bit.



Didn't realize this was out. Here's what I wrote in the other thread:

It's Gallup.

Inconsistency is their pattern. Don't forget this is the same polling unit that had McCain up by 10 points (when everyone else had Obama up or McCain up by low, low single digits) only to have McCain down by 7 in their next poll.

Averages folks, averages are the key. That + the elimination of automated (Rasmussen, PPP, etc.), internet (Zogby, YouGov, etc.) and daily polls (Gallup, Rasmussen) and you should be fine.

There's just far too much variation in Gallup's polling to take them seriously.



It would be nice to see the Demographic breakdown of this poll, because Melvin is right. The results of this poll implies approval among Whites at around 30% which seems very, very unlikely.



How is it that a USA Today/Gallup poll is 10-11 points off from what the regular Gallup poll shows?

Regarding demographics, Obama's been consistently averaging 38% approval among whites for the past four weeks in Gallup. 90+ among blacks and 60% among nonwhites. Unless they underpolled minorities, these results seem strange.

Anyways, out of all the pollsters Gallup seems to bounce around quite a bit, as can be seen weekly with their generic ballot.



"I can't believe how many poll deniers are out there."

Too much analysis hinges on the 1994 generic ballot. Republicans everywhere are frothing at the mouth for 1994 because that is the one win they've had in the past 62 years that would bring them to the kinds of gains they need this year.

Like I said yesterday evening, 60+ seat gains in one election simply. don't. happen. anymore. The parties/members are too entrenched, the voters too polarized.

In continuation of some of what we were discussing last night, I feel that the demise of the solid south for democrats and the closely divided partian nature of American politics since circa 1998 means that either party will have trouble making huge gains, and even if they do, holding more than 250 seats in the house for long.

It looks as if 235 is about as strong as either party can comfortably be given district partisan makeups, although the democrats seem to have about an upside a little higher than republicans (10-25 seats).

So we may be entering an era of lower seat turnover, but more switches of control because majorities are going to float +/- ~20 seats of that 218 mark. Haven't seen that for more than 100 years.



Aaron, you post does not address my two main points.

1)Do you believe in the magnitude of polls that show Repubs with a sizable lead?

2)If you do believe that GOP has the lead, do you believe it is possible for GOP to win the majority of popular House votes (in accordance to their lead in the polls) and still not win the majority of seats?

As far as what things do and "simply. don't. happen. anymore" there are many of those that are happening now.

1) For the first time ever, voters prefer someone else over incumbent.

2) Congress approval never this low.

3) GOP leading in Gallup generic. Never happened in Gallup's generic polling history. And history goes back to 1950's.

4) Huge enthusiasm gap.

Let's also remember that NJ, VA and Mass happened when BO's approval was in the 50's and the environment was much friendlier for Dems than it is now.

Bottom line, one can't just talk away polls. ALL of them? Polls are what they are.




"1)Do you believe in the magnitude of polls that show Repubs with a sizable lead?"

Yes, but I think the numbers are more indicative of people's true party inclination, less the "reagan dems" that always gave them a buffer.

"Congress approval never this low."

What is it, 12%? But individual congresmembers are significantly higher. Congress was at 14% approval in October 2006. Resulted in Dem +31 seats.

"it is possible for GOP to win the majority of popular House votes"

Yes, I just don't think the seat gain will exceed 50.

"voters prefer someone else over incumbent."

Grass is always greener and we are becoming increasingly cynical as a culture. I find that irrelevant since named republicans don't always do as well as generic ones.

" NJ, VA and Mass "

NJ and VA usually buck the national party in those off-year elections. These are overanalyzed. Did you ever WATCH those respective candidates? If you did, you'd see how impossible it would have been for Creigh Deeds to win. Guy talked like Foghorn Leghorn.

Coakley f**ed up big and there's a latent sexism in Mass (the whole country really, but more pronounced in MA) that gives a charismatic man an advantage over a woman. No female ever elected to Sen or Gov from Mass. Plus Scott Brown made a good argument saying MA residents would have to pay twice for health care.

Corzine...highly unpopular incumbent. I'm glad he lost. Charisma factor in play here too.



""1)Do you believe in the magnitude of polls that show Repubs with a sizable lead?"

Yes, but I think the numbers are more indicative of people's true party inclination, less the "reagan dems" that always gave them a buffer."

Not quite sure what you mean by that, but the question that's usually asked by generic ballot pollsters is something like "In your district, do you plan to vote for the Republican or the Democrat?" Pretty basic.

Today's poll averages point to something like 80+ seat gain. I don't really buy it, and have posted that I think the gap will narrow some. That's why I've stuck with my 60 seat gain prediction. So, since you're willing to go as high as 50, I say that's close enough.

I won't hold it against you when it turns out that you were off by mere 10-20 seats.

Seriously though, I predict mostly based on polls with a dash of momentum thrown in. I think GOP may be peaking too early right now. But I will adjust my estimates as the polls change. I was predicting 50-55 seats on this site a couple of months ago.


Chris V.:

What really scares me here is that only 25% of the people polled thought that it was "right" for WikiLeaks to leak the documents.

Did the people polled here fall for the pandering, inaccurate, fear-based narrative that leaking the documents jeopardizes our safety or the safety of troops in Afghanistan? I guess Obama is really not that much different than Bush after all. When someone releases information that reveals the ugly truth behind one of the wars Obama is waging, he bends backwards to vilify them and accuse them of being irresponsible and of endangering national security. When Bush did the same thing, us liberals called him disingenuous, and we were right. Ideological consistency requires us to do the same with Obama.

If the people polled had actually read what the leaks were about, they would know that most of what they revealed is that war is hell, that an awful lot of civilians die in war zones, and that (surprise!) an unstable Islamic government that we give a bunch of money to in the name of counterterrorism ends up promoting violence and terrorism against us anyway (ask 9/11 victims how well giving money to Saudi Arabia over the years to "fight terror" has worked out for us).



Post a comment

Please be patient while your comment posts - sometimes it takes a minute or two. To check your comment, please wait 60 seconds and click your browser's refresh button. Note that comments with three or more hyperlinks will be held for approval.