Tuesday, March 24, 2009

Arkansas Approval Numbers

Barack Obama is in positive approval territory in Arkansas, but just barely. PPP finds 47% of voters in the state expressing support for his work so far with 45% disapproving.

Like in most states there's strong polarization along party lines when it comes to Obama's job performance with 80% of Democrats but only 18% of Republicans approving. Obama also gets poor reviews from independents, with 54% disapproving and just 32% in approval. Those weak numbers with unaffiliated voters have become a trend in PPP's polls, including a national survey last week that showed his numbers with that group at 41/50.

Women and African Americans approve of Obama's work, while he gets negative numbers from whites and men.

One politician whose popularity is unambiguously high is Governor Mike Beebe. With an approval number of 68% and just 20% disapproving, he has the best ratings of any politician PPP has polled on in the entire country over the last year. Beebe even has a 55% approval rating with Republicans. He should be close to unbeatable for reelection next year.

Mark Pryor also benefits from solid numbers with 54% of voters in the state expressing support for his job performacne and 30% dissenting.

We'll release numbers looking at Blanche Lincoln's standing tomorrow.

Full results here.

13 comments:

Anonymous said...

Jensen, Jensen,

You never learn your lesson.. Your poll has 41/31/29 D/R/I whereas the November 4th exit poll showed 36/32/31 D/R/I.

You have black voters comprising 15% of the Arkansas electorate...they were 12% of the Arkansas electorate on November 4th. So you are projecting a 25% increase in black turnout from a year with historic black turnout. Hoping for ACORN a little too much, Jensen, eh?

I should just keep on schooling you but I'll just stop and let the commentators digest the data that I have cited and come to their own conclusions....

By the way, how much did you have Obama beating Hillary by in Pennsylvania in the April primary?

Brandon K said...

Any hint on what the Lincoln results show?

Eureka Springs, AR said...

Anyway we could see the actual wording of the questions... and the possible responses given?

Also as many as 20 percent (statewide) of voters cast a vote for Green candidates last November... why exclude 20 percent from consideration a few months later?

Tom Jensen said...

Gallup found a 12 point Democratic ID advantage in Arkansas last year. Exit polls are not the gospel truth. We don't weight for party id and that certainly seemed to serve us fine for our general election polling.

Anonymous said...

Tom,

Don't listen to these idiots. I think they are forgetting, that PPP was one of the most accurate pollster's in the Election. Thanks for all the hard work, that you all do. Proud to be in NC, with such an accurate pollster!

Anonymous said...

It all starts with the so called independents and in another month or so, obama will not have any support among that group. If this continues, it will be carter again

Anonymous said...

Jensen's logic seems to be Gallup is more accurate than exit polls, hahaha. Jensen, did you know where Gallup ranked on the list of twenty national pollsters?

http://www.fordham.edu/images/academics/graduate_schools/gsas/elections_and_campaign_/poll%20accuracy%20in%20the%202008%20presidential%20election.pdf

17 out of 20, hahaha. Gallup had Obama winning by double-digits, the final result wasn't even close to that. Then again, Frank Newport from Gallup shouldn't feel too bad as your boy Dean was quoted in a press release stating the following "Public Policy Polling’s final survey in Pennsylvania continues to show
an extremely tight race. Barack Obama has a 49-46 lead over Hillary Clinton in the state."

This is almost too easy, c'mon Jensen, ARG performed pretty darn well also and at least, it didn't have Obama beating Hillary in PA. Anyone with a map and an understanding of geography could tell that but thanks for your effort, hahahaha.

Rasmus said...

Anonymous,
you clearly have no understanding of political history.
Arkansas is one of the bluest states of the country, it beats Connecticut, Massachusetts and such states handily- I'd rather say that the PPP sample is a bit skewed towards the GOP and the Independents.
The problem for the Democrats is, that those voters are the old Dixiecrats- they didn't reregister after 1964 and a lot of them vote for the Republican in presidential elections and a straight Democratic ticket down on the ballot.
Wonder why the GOP didn't ran an opponent against Mark Pryor? Because he wouldn't have gotten 30%. Why is the AR house split 75 Democrats to 25 Republicans? Because the state is so damn Democratic.

So lets come to the exit poll. The exit poll asks for the Affiliation of a voter. You must imagine- the voter comes out of the precinct just having voted for McCain. So he is more likely to identify himself as a Republican or Independent.

Now, half a year later he gets called by a pollster and is asked for his party affiliation. He is a registered Democrat, so he answers Democratic.

By the way- Hillary Clinton got more primary votes than all GOP candidates combined- and the AR GOP primary was still competitive, with a favorite son candidate in the race.

Your next point- black voters comprising 15% of the AR electorate.
You cite the exit polls who comprised 12% of African-American voters. An exit poll is just that- an exit poll. The Arkansas exit poll had 1,321 respondants, what means that it has a margin of error of about 2.8 %. And this would assume that the poll was truly random. It isn't, since I'm sure the number of precincts is bigger than the number of respondants in the exit poll- whole precincts weren't exit polled.
The percentage of AR black voters could easily have been 14% without contradicting the exit poll. Also, you claim that 2008 was a year with historic black turnout. At the same moment, you write that Blacks made up 12% of the electorate in a state where 16% of the population is black.
That doesn't quite fit together, eh? The reason for that? No historic black turnout for you. The Obama campaign didn't contest Arkansas. No GOTV, no voter registration efforts, no competitive race results in low turnout.
This means there is still room to improve, and a 1-2% increase of the share of black voters doesn't seem so unlikely. We saw that happening in a lot of other states.
And you know what?
I hate to break it to you, but the 2004 exit poll in fact had blacks comprising which percentage of the electorate? Hm, can you guess it?
15%.

OK, on to your last point.
This is a quote from the paper you cite there:
[...]which is only about 1.37 percentage points away
from the current estimate of a 6.15-point Obama margin in the national popular vote.
_______________________
Now, sorry to destroy your nice little wingnut dreams, but your source is every bit as wrong as you are.
Obama won 52.87% of the popular vote, McCain 45.62%.
You know, that is not a 6.15 point margin. It is a 7.25 point margin.
And do you know what this further means? That your precious source, the only one that you have, bases its list on wrong election results.
Oops.

I should just keep on schooling you but I'll just stop here and let the commentators digest the data that I have cited and come to their own conclusions....

With regards to the PA Primary Poll- wow, PPP got one poll dead wrong in the whole election cycle. Really bad. Too bad that this also happened to every. single. other. pollster. Rasmussen, SUSA, whatever.

Do you have any arguments left?

Anonymous said...

Rasmus,

Too easy to shoot the crap you just posted out of a barrel.

Here's Gallup's final pre-election poll: http://www.gallup.com/poll/111703/Final-Presidential-Estimate-Obama-55-McCain-44.aspx

That looks like a double-digit victory for Obama against McCain, doesn't it (I'm assuming that you can read)? Oh wait, Obama didn't come close to an eleven point win against McCain. Since this poll is the only one on which we can judge for the national presidential election, Gallup doesn't seem to have too much credibility given how wrong it was with its final poll. One reason why Gallup may have been so off is that it incorrectly assessed the Democrat advantage in ID over Republicans.

I'll let you out of the classroom now Rasmus...it's public school, so you won't have to pay me for this tutorial.

Anonymous said...

Rasmus,

Also, find another pollster who all by himself missed a result like Jensen did with PA. All the pollsters missed NH but only Tom Jensen missed PA. Nobody missed PA except for Jensen. Find an analogous situation where someone allegedly of his firm's caliber missed an obvious result. Anyone who had a brain and a map knew that Obama was going to get embarrased in PA. Instead of just stating the proposition that everybody is wrong once in awhile, why not provide an example of where another respected pollster was as wrong with another state as Jensen was with PA?

Regarding Arkansas, have you considered for a second that Obama turned a ton of Arkansas Democrats into Republicans? Rasmus, I hope you and Jensen understand that not every white person who votes against Obama is a racist. You ever consider that these Southern White Democrats from Arkansas may no longer consider themselves Democrats because of Barack Obama?

By the way, ARG nailed the election.

http://americanresearchgroup.com/archives/?p=23

"Our final weekly national survey had Barack Obama at 53% and John McCain at 45% (scroll down for details).

Our final electoral vote estimate had Barack Obama winning 28 states with 353 electoral votes and John McCain winning 21 states with 163 electoral votes. Indiana and Missouri were tied in our final surveys. We did not have a large enough sample size to predict Nebraska by Congressional District."

At least ARG had enough sense to show McCain winning Montana. Any other final pre-election poll showing Obama winning Montana other than your boy Jensen?

I should really charge for classes.

Rasmus said...

Anonymous,
I looked at this one-
http://www.gallup.com/poll/111664/Gallup-Daily-Obama-Continues-Outpace-McCain.aspx

I guess Gallup is really a bit insane, since they'd must have had Obama up by something like 20 points on their last tracking poll day to influence a ROLLING average like that.
We don't really argue about Gallup though- we argue about a PPP poll.

"

Regarding Arkansas, have you considered for a second that Obama turned a ton of Arkansas Democrats into Republicans? Rasmus, I hope you and Jensen understand that not every white person who votes against Obama is a racist. You ever consider that these Southern White Democrats from Arkansas may no longer consider themselves Democrats because of Barack Obama?"

I do. If this was not the case, I would complain that the sample was way too much leaning Republican. In the pre 9/11 era, Democrats had a party ID advantage of something closer to 20-25%. That has obviously diminished in the Kerry-Pelosi-Obama era. But there still IS a Democratic Party Identification Advantage, as this poll (PPP doesn't weight for party, so those are the raw number) and the Gallup poll.

There is another way to look at this- the 2006 exit polls, when there was no upticket race.
Unfortunately, there was no statewide election in Arkansas, and House exit polls don't exist.
But we can look at the West Virginia exit polls, a state that has somewhat similar dynamics.
There, Democrats had a 19-percent Party Identification advantage. In 2008, that closed to 14-percent.
The Obama effect seems to have shifted the Party ID in West Virginia by 5 percent. Will that keep up for midterm elections? I don't think so, but it will in the 2012 elections. Byrd and Rockefeller can still count on the support of a democratic MAJORITY, not just plurality, while the national candidates can not.
If we assume- and I know, that is an uncertain assumption, but we simply don't have better data on Arkansas- that the same 5-percent shift would hold up for AR, we'd have a 9-point Democratic PartyID advantage for midterms, compared to a 4 point-advantage for Presidential Elections.
It would be easier if there had been exit polls on the Pryor race.


Second- what about ARGs Iowa Caucus poll? Clinton up 34-25 over Obama with Edwards at 21.
Or something like the SurveyUSA Indiana primary poll? Clinton +12, she won by 1. And no other pollster was that far off, closest to SUSA was ARG at Clinton +8.
Zogby showing a tie in OH before the primary? Suffolk and ARG showing Obama leading in Oregon by just 4-5 points before the primary (you know, he won by something like 18. Oh, and PPP nailed it).

In PA-Primary, your beloved ARG had Clinton up by 16- almost as far off as PPP.
In the Wisconsin primary, ARG showing a 16 point shift towards Obama in two days, still underestimating him by 8?
ARG showing a 3-point Obama lead before the SC-primary?

Primaries are notoriously hard to poll, because of the low turnout. Mistakes happen, there.
Oh, and read that:
http://publicpolicypolling.blogspot.com/2008/04/breaking-down-our-pennsylvania-poll.html

Can you imagine ANY OTHER POLLSTER doing this? No? That's why PPP is so popular.
ARG put out a boatload of shitty polls this year- they're one of the worst, together with Zogby.
And PPPs Montana poll was actually within the margin of error.

Rasmus said...

Actually Gallups PartyID advantage for Democrats wasn't that off, actually. It's that they had way too many independents and less partisans-and those independents were breaking for Obama in their poll. Not in reality.

Rasmus said...

Oh, and by the way- your main point was that the racial composition of the poll was wrong. This took also up the most space in my response. So...why did you drop this, hm?

 
Web Statistics