Pollster.com

Articles and Analysis

 

POLL: ARG WI, SurveyUSA TX


American Research Group

Wisconsin
Obama 52, Clinton 42... McCain 51, Huckabee 43, Paul 3

---
SurveyUSA

Texas
Clinton 50, Obama 45... McCain 50, Huckabee 37, Paul 7

 

Comments
Henry:

Amazing new numbers in this Gallup poll, showing Obama's gains nationally since Feb 5 amongst literally every subgroup:

http://www.gallup.com/poll/104428/Obama-Gaining-Among-MiddleAged-Women-Hispanics.aspx

Most striking to me on first look is that compared to two weeks ago Obama is now leading among Hispanics, as well as "core" Dem voters, and the middle age group.

Really interesting to compare these numbers. The Clinton campaign has got to be really worried.

Henry

____________________

polsaa:

The new ARG poll is pretty funny. Apparently they looked around, saw that they were on the opposite side from everyone else, and jiggered their numbers to put Obama ahead. Now they won't look like idiots again if Obama wins.

____________________

Greg:

Seriously, should ARG polls even be reported? They clearly have no idea what they are doing.

____________________

Dave S:

So much for looking at trends among a single pollster! Is ARG good for *anything* other than providing extra plotting points for pollster.com?

____________________

Polsaa, I agree 100%. It looks like they're trying a backdoor path to respectability. Either that or the amount of effect weekend polling has vs. weekday polling is HUGE.

Also, there's a new pair of Senate polls out from Rasmussen (OR, MN).

____________________

Caden:

Hey ARG, where can I short your stock?

____________________

Shane:

No comments on the SurveyUSA poll? It seems pretty big to me....

With the way delegates are given out and with the caucus aspect.. I basically can't see how Clinton can win Texas at this point. Sure, she might be able to spin a slight popular vote win into a victory, but that's not what she needs. She needs delegates.

____________________

Adam G:

Not that I'm a fan of ARG, but in their defense, their initial poll in WI was performed just after Super Tuesday, before Obama's big run of victories. Believe it or not, both of their polls could have been correct.

____________________

rowjimmy:

"I basically can't see how Clinton can win Texas at this point."

Is the election held today at +5 or in two weeks at a different number?

She needs delegates, but she also needs to change the tune in the papers about "winning." The majority of people still do not understand "delegates" let alone that there is anything called a "superdelegate." Accordingly, even if she wins Texas by just 2% and (assuming) Ohio by 6, the headline reads the next day, "Hillary wins Texas and Ohio!"

That changes the game somewhat. Obama's camp will do the "she cannot catch us. Game. set. match," as you are saying here. But, if by some miracle (I don't see it happening either) she wins both big states on March 4, do you REALLY think that the front page headlines will read on the morning of March 5, "Hillary still far behind in Delegates" followed by a story on A14, "Hillary wins Texas and Ohio; set to drop out tomorrow." Is this the scenario you are painting? It seems like you are.

Texas should be close. If either candidate wins by more than 6, I would be surprised. I think the white Democrat vote is too high for her down there. Also think that hispanics will be more than 32%, but that the MOV among African-Americans will be larger than even SUSA shows.

JMO.

____________________

Adam G:

In fact, ARG talks about it on their website:
http://americanresearchgroup.com/

Poll #1: Feb 6-7 (though not reported on pollster until last weekend I believe)
Poll #2: Feb 17-18

Two weeks is a political lifetime...

____________________

rowjimmy:

"Believe it or not, both of their polls could have been correct."

So you think there was a 20 point swing among women in 1-2 days from Feb 15-16 to Feb 17-18? And a 16 point swing overall?

I think they are trying to save face by cooking some numbers to reflect other polls. No way did the electorate, no matter HOW volatile, change by 15-20 points in 1-2 days. Even if Hillary gave the crowd the finger and Bill went Hannibal Lecter on someone, I don't think you'd see that kind of swing. It's purely a face saving move on ARG's part. Or they had extremely incompetent polling on Feb 15-16.

____________________

jack:

For you information, Survey USA is the most reliable poll you jackass so think before you speak again about Obama winning Texas.

____________________

Adam G:

rowjimmy -- read my comments above.

____________________

rowjimmy:

"Poll #1: Feb 6-7 (though not reported on pollster until last weekend I believe)
Poll #2: Feb 17-18

Two weeks is a political lifetime..."


Adam G,

You miss the poll on Feb 15-16 from ARG. There were three polls, not two. I could buy the change over two weeks.

The ARG poll went from 49-43 to 42-52 in 1-2 days... or are you still unable to see this?

____________________

rowjimmy:

"rowjimmy -- read my comments above."


Adam G,

There were THREE polls...Not two. Not the huge switch in 1-2 days. Not 2 weeks.

Read poll results below, from ARG:

Democrats Feb 6-7 Feb 15-16 Feb 17-18

Clinton 50% 49% 42%
Obama 41% 43% 52%
Someone else 1% 1% 1%
Undecided 8% 7% 5%

____________________

Adam G:

rowjimmy -- As for the second poll, I certainly stand corrected! I knew of only the two.

It is possible (given a very volatile environment) for the swing to happen in a two day period. But with all of the other polls taken over the same period conflicting with ARG? Yeah I don't buy it...

____________________

Shane:

"That changes the game somewhat. Obama's camp will do the "she cannot catch us. Game. set. match," as you are saying here. But, if by some miracle (I don't see it happening either) she wins both big states on March 4, do you REALLY think that the front page headlines will read on the morning of March 5, "Hillary still far behind in Delegates" followed by a story on A14, "Hillary wins Texas and Ohio; set to drop out tomorrow." Is this the scenario you are painting? It seems like you are."

You don't seem to understand. The math is against Clinton. For example, an analysis of a poll showing Clinton with an 8% LEAD still gave Obama 8 MORE DELEGATES because of the way delegates are given out. And thats BEFORE counting the caucus portion of Texas! Caucuses always go for Obama (don't let New Mexico confuse you, as far as I can tell, it wasn't really a caucus), so I just don't see the math that leads to a true Clinton victory in Texas.


"For you information, Survey USA is the most reliable poll you jackass so think before you speak again about Obama winning Texas."

I don't know who this was directed at or who you even think will win in Texas, but, if it was directed at me, please consider my prior words.

____________________

rowjimmy:

Adam G,

I can see where you came to your conclusions. In their text "summary" of the polling data, they seem to overlook their 2nd poll on Feb 15-16. They try to paint a picture of "momentum" from Feb 6-7 to Feb 17-18, when their agency actually showed a nearly similar poll on Feb 15-16 to that taken on Feb 6-7. So, in ARG's worldview, momentum expresses itself in 24-48 hours, when there was a 20 percent exodus of women from HRC's core support.

Maybe momentum is non-linear in ARG's polling? Or maybe someone at FEMA took the Feb 15-16 poll? :-)

____________________

Joshua:

ARG... Damn, now you're just being crazy.

____________________

Jordan:


I enjoy the ARG numbers. I see them as entertainment value. Gives us something to shake our heads at, complain about, critic, or in my case: chuckle.

"You are never COMPLETELY worthless, you can always serve as a bad example."

I think it also reminds us to take EVERY poll with a grain of salt.

____________________

Joshua:

In Clinton's defense, I do remember Obama getting a "surge" in support a couple days before big Super Tuesday races such as California, Massachusetts, and New Jersey. That didn't work out too well for him.

Obama surges seem to taper out come election day.

____________________

Joshua:

The Feb. 17-18 ARG poll is the first time I've ever seen ARG feel the need to explain themselves.

Simply amazing.

____________________

rowjimmy:

Adam G:

You are talking about math. I get the math. Trust me.

However, I am talking about headlines on the front page should Hillary pull off two big-state wins on March 4th. Those would NOT be about "delegates." The talking heads and pundits would surely talk about that following the results. But the immediate impact would be to change the election landscape since Obama was inevitable and Hillary dead; then all of a sudden, she wins two huge states. Yes, she would still be lagging behind in pledged delegates (though, how "pledged" they really are is another matter all together).

They are different subjects...math versus headlines on March 5.

Will probably be moot though. I do not expect Hillary's so-called "firewall" to hold in Texas.

____________________

rowjimmy:

"The Feb. 17-18 ARG poll is the first time I've ever seen ARG feel the need to explain themselves."

That they had to "explain" themselves speaks volumes. Sort of like Zogby and his absurd +13 in CA. ARG's explanation reads about as convincing as Roger Clemens' testimony to congress...

:-)

____________________

Henry Rosen:

Overnight IVR Poll shows Obama Leading Clinton among Strong Democrats - households that have a consistent pattern of voting in Texas Democratic primary

http://texaspolls.callvoyager.com

Obama - 40%
Clinton - 33%
Undecided - 14%
Otherwide - 13%

These IVR polls are proving to be very accurate - they are overcoming the changing dynamics of live polls; the sample size on this one is 1,600+ with a MoE of only 2.4%

____________________

rowjimmy:

"Overnight IVR Poll shows Obama Leading Clinton among Strong Democrats - households that have a consistent pattern of voting in Texas Democratic primary 40-33"

Wow, that's very different from SUSA, which shows Clinton leads 55-41 among Democrats (78% of expected primary turnout).

____________________

rowjimmy:

"These IVR polls are proving to be very accurate"

Henry, by what metric do you assess this claim? I am curious.

____________________

Joshua:

"That they had to 'explain' themselves speaks volumes. Sort of like Zogby and his absurd +13 in CA. ARG's explanation reads about as convincing as Roger Clemens' testimony to congress..."

This primary election season has taught me a new rule in regards to opinion polls: If it sounds unbelievable, then it's probably not true. It seems to be a consistent rule so far.

____________________

roy:

Shane, the SUSA poll was discussed in last night's CNN TX poll thread. The crosstabs are amazingly extreme--I don't think the gender gap has been _that_ large elsewhere. A couple of things that struck me:

Clinton dominates among Latinos but not among people for whom immigration is the most important issue. I honestly don't understand the basis for the overwhelming support...I would have expected them to split the Latino vote.

Clinton runs strongest among Democrats (btw, 11% Independents seems rather low), while Obama runs strongest among liberals. I understand that not every Democrat is a liberal, but it's still weird.

____________________

Joshua:

Also, maybe this makes sense to well-informed people, by why is ARG using a Bible verse (Mark 4:25) in its explanation of its Feb. 17-18 poll?

That strikes me as crazy. You don't read much Holy Writ when you're reading through opinion poll results.

____________________

The math is the math, and we can talk about methods all we want. The simple reality is that primaries have not been kind to pollsters for a number of reasons, probably most important among them being the fluid nature of voters (they are making their minds up at the last minute) and open primaries being very difficult to call "likely voters" on.

Having said that, the way the polls are used in the press is a different reality altogether. The worst case scenario for Clinton is a small Obama lead in Wisconsin. On the front end, the story is that Obama is ahead, but on the back end, the inability to find independents and young people gives him a chance to blow out expectations.

That's a lot of what we've been seeing lately, and the pollsters apparently want to correct their math to be more predictive. I would guess it's a point of pride, but I would bet that they don't like being a total tool for the press. The last ARG poll in Wisconsin was pumped full of more fudge than an eight year old at the State Fair.

I don't think that the crosstabs are going to prove as useful as we'd like as these polls are analyzed and adjusted. Once again, the situation is very fluid. Democrats tend to like both candidates but are unsure of just how they want to make history. At this point, you'll see a lot of momentum to Obama just to get it over with, and that's probably the only strong motivating force that you'll find.

In short, I think that Wisconsin has completely played into Obama's hands, and Texas will as well as we go forward. Clinton simply hasn't done anything dramatic to stop the momentum at this stage, and Wisconsin will only accelerate things.

The math? It's become a plaything for the pundits, sadly. But it's not like that hasn't happened before.

____________________

Adam G:

Erik Hare -- What are you saying that Clinton hasn't done anything to stop the momentum!?!?!?
Yesterday the Clinton camp showed that Barack is nothing but a plagarist! LOL.

Seriously though, if that's all they've got, they're getting really desperate.

____________________

Another Mike:

"For example, an analysis of a poll showing Clinton with an 8% LEAD still gave Obama 8 MORE DELEGATES because of the way delegates are given out. And thats BEFORE counting the caucus portion of Texas! "

Please do not think that one particular scenario is how it must be. First, that analysis was wrong because it allocated all undecideds to Obama. Second, it really depends how the votes in the particular senate districts come out. Just like it's possible for Clinton to win the popular vote and lost the delegates, it's also possible for Obama to win the popular vote and lose the delegates.

____________________

Shane:

"Please do not think that one particular scenario is how it must be. First, that analysis was wrong because it allocated all undecideds to Obama. Second, it really depends how the votes in the particular senate districts come out. Just like it's possible for Clinton to win the popular vote and lost the delegates, it's also possible for Obama to win the popular vote and lose the delegates."

Yeah, your last statement is true, but the real issue is this: Hispanic votes will be undervalued in terms of delegates and African-American votes will be overvalued. Its still possible for Obama to win the popular vote and lose on delegates, but it would take some sort of miracle.

____________________

Another Mike:

"Clinton runs strongest among Democrats (btw, 11% Independents seems rather low), while Obama runs strongest among liberals. I understand that not every Democrat is a liberal, but it's still weird."

This actually fits with the other poll stating that Obama does better with strong Deomcrats. For whatever reason, Obama seems to flank Clinton--he wins with liberal Democrats and independants; she wins with moderate and conservative Democrats (which are a pretty big chunk of the electorate). At least that's one plausible theory to explain the different information. Or, some of the polls could just be wrong!

____________________

Shane:

"This actually fits with the other poll stating that Obama does better with strong Deomcrats. For whatever reason, Obama seems to flank Clinton--he wins with liberal Democrats and independants; she wins with moderate and conservative Democrats (which are a pretty big chunk of the electorate). At least that's one plausible theory to explain the different information. Or, some of the polls could just be wrong!"

It seems to me that Obama definitely flanks Clinton. My friends who have voted Green in the past will vote for Obama in a GE, but they would never vote for Clinton. Same with some of my evangelical Republican friends, my moderate anti-war friends, and my libertarian friends. Out of old school democrats, though, there is certainly a greater division. Within my dad's family, for example, there is more mourning over Edwards than support for Clinton or Obama.

____________________

Anonymous:

"Hispanic votes will be undervalued in terms of delegates and African-American votes will be overvalued. Its still possible for Obama to win the popular vote and lose on delegates, but it would take some sort of miracle."

That depends. If Hispanic dominated Senate Districts increase their share of the statewide vote, then they will be undervalued. This seems likely. If African American dominated Senate Districts decrease their share of the statewide vote, then they will be overvalued. I'm skeptical this will happen.

My real point is that I think the Obama-has-structural-advantages in the Texas primary theme has been way overplayed. I'm afraid Obama is losing the expectations game on this issue.

____________________

New SUSA polls - Clinton lead down to 9 in Ohio, Barack beats McCain in Iowa and Virginia while Clinton loses both states.

____________________

rowjimmy:

"Clinton lead down to 9 in Ohio"

You can only say it's "down to 9" if they (SUSA) had previous polls that showed here up by more. I believe the other polls showing her +13 or whatever were from different agencies.

I'd say that looks good for her. If she can be outside the MOE after 2 weeks of ridiculously negative coverage about her campaign, it's not a bad sign.

A terrible showing today would change much of that though. If she can lose by around 5ish, she will claim she beat expectations and try to recover in the debates.

____________________

Another Mike:

"You can only say it's "down to 9" if they (SUSA) had previous polls that showed here up by more. I believe the other polls showing her +13 or whatever were from different agencies."

SUSA did have a prior Ohio poll taken 2/10-11. In that poll, Clinton led 56-39.

____________________

NHCt:

ARG just came out with a poll with Obama leading in Virginia and Maryland.

____________________

Joshua:

Here's another way of describing the ARG Wisconsin poll that might be enlightening to people.

So far this primary season, I have seen a situation develop where two pollsters have radically different results, and it's apparent that one pollster is going to get proven right and the other is going to get proven wrong. An example of this would be the California polling where Reuters/C-SPAN/Zogby had Obama leading Clinton by an amazing percentage and SurveyUSA had Clinton leading Obama by an amazing percentage.

However, with ARG now, this is the first time I've seen a single pollster set themselves up where one of their own polls is going to be proven right and another of their own polls is going to get proven wrong. Which is going to be right? The Feb. 15-16 ARG poll that shows Clinton ahead or the Feb. 17-18 ARG poll that shows Obama ahead?

____________________

rowjimmy:

@Another Mike:

I realized my SUSA mistake shortly after I posted it. Sorry. I try not to make such mistakes. With so many polls out there, I just overlooked it.

It will be interesting to see what, if any, effect the debates and spin from recent stories have on Texas and Ohio.

____________________



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.

MAP - US, AL, AK, AZ, AR, CA, CO, CT, DE, FL, GA, HI, ID, IL, IN, IA, KS, KY, LA, ME, MD, MA, MI, MN, MS, MO, MT, NE, NV, NH, NJ, NM, NY, NC, ND, OH, OK, OR, PA, RI, SC, SD, TN, TX, UT, VT, VA, WA, WV, WI, WY, PR