Pollster.com

Articles and Analysis

 

POLL: Zogby IN, NC (5/1-2)


Zogby Tracking (release)

Indiana
Obama 43, Clinton 42

North Carolina
Obama 46, Clinton 37

 

Comments
jac13:

I find the IN results more interesting than NC; although NC appears to be tightening, I still think it would be difficult for Clinton to make up 9 points in two days.

As for IN, it could just be statistical noise, but it looks like Obama picked up a point from the undecideds. If that turns out to be the leading edge of a trend, it would be a reversal of the pattern in the past few primaries, where the undecideds broke for Hillary. Am I right that, even if they break equally for both candidates, that's a boon for Obama?

____________________

Dan:

I've said this elsewhere, but I really think that Indiana has the potential to be much closer than people are expecting. The polls have see-sawed back and forth; moreover, only Howey-Gauge (based in Indianapolis) has an history of polling the state (suggesting, perhaps, that many of the other pollsters are doing a fair bit of guesswork about demographics).

I won't make any predictions, but IN has the potential to be a nail-biter, like Missouri. (Remember that SUSA had Clinton up by 11 points that day before the Missouri primary.)

Guess we'll find out soon enough.

____________________

jac13:

Dan,

Not soon enough to suit me! I'm on pins and needles about these two -- there's a lot riding on it for Obama (whom I support). He'll be OK if he loses IN close and wins NC (at this point, by ANY margin).

I used to fear an IN blowout, but for the past couple of weeks I've been thinking, like you, that whoever wins it will do so by a few points, and the polls seem to bear that out. There are some encouraging signs for Obama in the early voting -- heavy turnouts in "his" counties.

If Obama can hang on in NC and eke it out in IN, as everybody has said, HRC is probably finished. However, if she wins NC it's a new ballgame.

____________________

richard pollara:

One shouldn't forget Zogby's stellar "day of the election" poll in California either. After watching these polls pretty closely for four months I am convinced that the daily horoscope has a better chance of fortelling the future than a slew of primary polls. The only things that really seem to matter in the primaries is the makeup of the electorate and who is controling the news cycle the last few days before the election. My sense is that in a toss up state like Indiana the momemtum is with Mrs. Clinton. But I can't confirm that until I check with my astrologer. As Dan said, we will find out soon enough.

____________________

jac13:

richard,

I'll grant you that Zogby laid an egg big time in CA -- could it have been the early voting?

Anyway, on the other side, he nailed the results in PA, both the late trend to HRC and her ultimate margin.

Tick tock.

____________________

desirous:

Wasn't the California poll Zogby Interactive? ZI is a veritable crapshoot. The regular Zogby tracking poll is mediocre but not THAT bad. After all, the tracking poll nailed Pennsylvania spot-on.

____________________

Jacob S:

All of you have made good points. I would note that this poll has a lot undecideds. As we saw Pennsylvania and Ohio, Obama tends to remain in the low 40s and Clinton tends to vary on the basis of how hard the pollsters pressure undecideds. I would look to see if Clinton begins to open a lead in Indiana in the final days. If so, here we go again with Ohio and Pennsylvania.

____________________

jsh1120:

Turned on my radio this morning to Weekend Edition on NPR and happened to come in part way through an unusually perceptive and balanced interview with a pollster discussing the state of the national race. How refreshing, I thought. Wonder who this guy is.

At the end of the interview I found out. Mark Blumenthal. Not surprising.

____________________

akamrearl:

I hope NPR's weekend morning shows continue to use Mark. he was great.

____________________

Uri:

I agree that IN may very well go to Obama (NC certainly will) and that in most cases HRC is finished. Even if not, she would be once the superdelegates rush in to fill the 100 person gap. You've seen how pleased with himself the guy who switched this weekend was in his big press conference. I bet there's tons of undeclared supers waiting to be #100.

As for the polls though, I think Zogby is reliable, but I've noticed even in PA that they generally poll HRC lower than BHM. Any idea why that is?

____________________

jac13:

Uri,

Put another way, they poll BHO higher than he actually performs, right? Could it be "the Bradley effect?"

____________________

killias2:

"Put another way, they poll BHO higher than he actually performs, right? Could it be "the Bradley effect?""

Except for, you know, the plethora of states where Obama did better than expected. Wisconsin anybody? The final polls were right on in Texas, Ohio, and PA, yet everyone screams Bradley Effect. What gives?

One thing I'll say about PA, people who made their decision in the last 24 hours went for Clinton 60-40. That makes polls like Quinnipiac (which had Clinton up by 7 with 6 percent undecided) pretty much spot on in that contest.

____________________

jac13:

killias,

Didn't the same thing happen with the undecideds in TX (primary, that is), OH and CA?

Obama's 1-point uptick in the Zogby IN poll appears to have come from the undecideds. It's probably statistically insignificant, but I'm hoping it portends a blunting of Hillary's usual late-deciders advantage. As I said earlier, even if the undecideds break equally for the candidates in IN, that would be an improvement for Obama, no?

____________________

carl29:

Some people talk about "Bradley Effect" and don't even know what they are talking about. The "Bradley Effect" is when people tell pollsters that they will vote for the African American candidate and when they go to vote don't do it. This has not being the case with Obama. If you go back to all the polls in PA they were right on target regarding his support, so no one lied to the pollster by telling that they will vote for him and then not following through. I and my husband, a Math PhD from Tufts University, paid attention to the polls in PA and noticed that Obama's support was always low 40's, he got 45.4% of the vote at the end. The only number that fluctuated was Hillary's. So people were between voting for her or not voting at all. They never lied to the pollster by telling they will vote for Obama.

I bet the same thing is going on with her in NC. Look at the numbers, the is stuck at low 40's. The number that changes is Obama's but it doesn't translate into more support for her. I suspect that low 40's is her ceilings, just like it was his in PA.

____________________

jac13:

carl,

Hope you're correct about NC.

The "Bradley effect" in PA -- if there was one -- showed up in the exit polls, where Obama apparently polled 4 to 5 points better than the percentage of the votes actually ended up with. Hence the media's initial "too close to call" reports.

____________________

Jordan Davies:

Zogby looks good and probably accurate. Indiana will be close, especially with a lot of early voting in districts favorable to Obama. NC should be easy, so we will see.

____________________

ByTheNumbers:

"Late deciders" are shaping up to be the last refuge of the polling scoundrel. If you blow the poll sample, all you have to do is say "there were a lot of late deciders going the other way."

Keep in mind that the only measure of late decider trends is yet another poll: exit poll data collected from people who are willing to fill out a lengthy form after they have voted. This is not a valid methodology for measuring the actual electorate; it captures only the subgroup that wants to feel like they are "participating" in something like a poll, which is not a random sample. It skews young, amongst other biases.

Even if you accept a 60-40 split of the "undecideds," that 20% gap only explains a 1 point swing for every 5 points of undecideds in the poll. So Rasmussen in PA, with 7% undecideds, only picks up 1.2% of swing on the late decider theory.

The real issues that are being highlighted by this year's polling extravaganza, in my view, are polling technique and weighting. Polls that use lengthy telephone interviews (30-40 questions, with minimum response numbers of 25+ questions) are tending to miss the mark (usually favoring Obama). The Downs Center and Bloomberg polls in IN fit this category. Other polls have shown significant skew based on weighting assumptions. The Zogby v. SurveyUSA gap in CA was caused primarily by weighting issues (Zogby blew the weighting badly because of faulty assumptions about turnout in certain sub-groups).

Both issues highlight the fact that this election features unusually high demographic separation: different sub-groups are voting very distinctly (more so than normal). And this is an often-hidden weakness in present polling: many polls do not have a truly representative sample because the polling technique inherently skews the responder set, and the pollsters are not controlling for this skew.

The question of controlling for responder skew in an election with significant demographic separation is a really interesting technical issue this year, and it should get more attention. Instead, the news media behave as though the polls are always accurate within their stated "margin of error", and the pollsters explain away their mistakes with the great "late decider" excuse.

Oh well.

____________________

carl29:

I think that the problem with the exit polls has to do more with gathering more information from rural areas, that are usually not covered that much, and more urban areas. These two candidates, Hillary and Obama, have different base of support that if you don't cover a good sample of "bases" the numbers are not going to be accurate. He does well in urban areas as we know; however, does poorly in rural areas, so if the pollsters don't send enough people to these rural towns they information is gonna be off. In PA for example, she sweeped small towns, some of the almost by 70%, but if how to know that if there were no pollsters getting the information.

My suggestion to those companies will be to send more people to small towns so they will know by how much she won those towns and weigh that in with information coming from more urban areas.

I remember that on Super Tuesday, the Associated Press made a fool of themselves calling Missouri for Hillary. In Missouri Hillary won big in rural towns, which are the first votes in; however, when the votes from Kansas City and Saint Louis came in, Obama support was so overwhelming that redressed the advantage Hillary had more the rural area. The associated Press apologized saying that they thought that there was no way for Obama to catch up, but he did.

____________________

skibum1981:

There was no Bradley effect in PA. Regarding the exit polls point, Obama voters notoriously are very enthusiastic voters, and thus are significantly more likely to participate in the exit polls. As far as I know, most states have seen Obama perform worse in the actual vote count than in the exit polls, with some notable exceptions such as MS.

____________________

boskop:

great reading! very informative to say the least. i'm no expert like the rest of you folks but i tend to go with macro 'feelings' probably the most nauseating concept for pollster people.

my gut tells me that hillary will take indiana rather comfortably. it is after all, a republican state and she does well in the heart land where more men poll in her favor.

nc is the most fun. my macro stomach tells me that maya angelou is not anathematic to the kind of trends you'll pick up. more middle class educated black women will leave the pack and vote for hc because they always wanted to
endorse a role model for their children, but needed some trigger. now they have one to pull they excuses them from the pressure of racial polarity and permits them to move over to the other side.

with wright and ayers, this is the trigger they needed. i think obama will win, but it will be game changer as the black vote will start to erode beneath him. macro vs micro [:>)

____________________

RS:

NC:
Zogby says the African-American vote is 73-10 in Senator Obama's favor. I think it will end up 85-15; given African-Americans are ~1/3 of the NC electorate, that means Senator Obama tops 50% (assuming the 46% hold.) That leaves ~11% undecideds/others; give that 2:1 to Senator Clinton. Overall, undecideds break 9-8 for Senator Clinton - that's because of the large African-American undecided vote. Finally, we have a 54-46 split in Senator Obama's favor.
As an Obama supporter, I'll take that happily.

IN:
Unfortunately, Zogby doesn't seem to give demographic breakdowns (which also applies to the NC results). Anyway - that +1% Obama gain is well within noise...

Damn pollsters who don't give cross-tabs... :-)

____________________

akamrearl:

Maya Angelou supported Clarence Thomas for the Supreme Court. She reasoned that the would come to his senses one day.

Don't expect a game changing impact on blacks. I think that community sees her as a skillful, lovable poet, but politically naive. That said, she may swing a few votes among non-black 'leaners.'

____________________

RS:

@boskop:
That commentary about NC is pretty heavily tilted toward Senator Clinton (but of course!)

You really believe that tangential Ayers connection will affect African-American (or even other) voters? Seriously? [As for Rev. Wright, a bunch of Black Chicago Ministers came out in support of Senator Obama.]

And you are assuming Maya Angelou or Senator Clinton will be a better role model than... ummm, the first African-American President of the United States of America.

Yeah, right.

____________________

carl29:

I have the gut feeling that she is going to take Indiana by around the same margin in PA. However, NC is Obama for sure and AA will back him at the same rate 90%. If I were part of the Obama camp, I will be all over the networks "crying" that senator Clinton is positioned to take NC away from us, raising expectations for her. So, when the numbers come in on Tuesday, the pundits have to give Sen. Obama credit for taking NC double-digits.

Look, AA are in the tank for Obama, they are supporting him because he is whom he is. No one can change that, sorry Maya. My point about raising expectations for her is to help him with perceptions. Every time that a state in the South with large AA community votes and Obama wins big, no one gives him credit. I remember when Mississipi voted right after Ohio and Texas. The polls showed that Hillary was "closing in", I even remember one that had her 4 or 5 points behind Obama. The spread according to RealClearPolitics was around 10%, so every one was talking about the big Mo Hillary had. The day came, Mississipi voted and Obama won by 25%. Hillary got crushed by the AA tide that backed Obama by 91 or 93%. The pundits took Sen. Obama's win from granted.

____________________

Tzal:

Is it fair to say that this poll shows the effect (or lack) of the gas tax debate between the two candidates? I haven't found any polling specifically on this issue and am curious about its potential impact.

Most of the recent polls were taken before this debate heated up, whereas this one was taken as the arguments developed.

____________________

akamrearl:

Dear Carl29: A year ago most leading reporters just knew that African-Americans were 'in the tank' for the Clintons. What's fun about this election - for me - is its surprises. I can't wait to see who next makes the mistake of taking any voting group for granted...

____________________

carl29:

The only thing that these polls show is that Hillary is stuck around low 40's, just as Obama was in PA. The only number that changes is Obama's. I think that her support is low 40's because AA's make up around 35% of the electorate and Obama gets 90% of their support, combined with 1/3 of the white support. I think among white he is around 35%. All her support comes from whites. In addition, too bad for her that Republicans cannot vote in NC's Democratic Primary, just independents, whom I suspect are less racists. Independents tend to be more socially progressive and more fiscal conservative, oh and tend to be more men than women.

____________________

carl29:

I understand that AA did not give Obama a chance. Let's be honest, their history is far from Optimistic. However, I think that Iowa and the other state that came after gave them hope that it could be possible this time. And let's not forget about Bill and his big mouth, trying to diminish Obama. All that put together encouraged AA to show up at the polls and made their voices heard. I remembered that I was listening to Ed Schultz when he said that everything Bill was saying would backfire at the polls. Up to that point AA were not sure if Obama was "black enough," but once Bill started attacking him they went sort of: What a minute, they are picking on a brother ! ! ! From then on AA embraced Sen. Obama as one of them.

____________________

boskop:

RS:

there you go again!

my statement stands. and btw: a black president of and by itself means absolutely nothing if he screws up!! we vote based on whether or not we think he or she will mess up faster. i believe barack obama should he become president will engage us more deeply in warfare worldwide than any president of the past fifty years.

as bush one said, "read my lips, no new taxes.." and found that he needed to change the game when he reached office, mr obama has NO IDEA what the ramifications are out there nor the possibility of his being tested.

in other words, his "read my lips, no new wars.." will come back to bite him simply because.

i would rather he not be there for that event. i would rather it be someone else. that is my right is it not? i lack however, the antipathy for obama people that you have towards hillary people. i find it very annoying and unfortunate to have to deal with reading this kind of sniping from you over and over on this site. sorry RS...

____________________

Paul:

As to Zogby and crosstabs, subscribers have access to detailed poll frequencies and crosstabs. Based on gender and race, here is what you need to know.

Based on the May 3 release, which was based on rolling poll over May 1-2, the poll frequencies and results in IN were as follows: overall HRC 41.7%, Obama 42.5%, someone else 7.4%, not sure 8.4%; Female 55% (HRC 43.6%, Obama 37.2%), Male 45% (HRC 39.4%, Obama 49%), White 83% (HRC 47%, Obama 37.2%), AA 11% (HRC 10.2%, Obama 81%).

Results in NC were as follows: overall HRC 37.2%, Obama 46%, someone else 8.2%, not sure 8.6%; Female 55% (HRC 39.2%, Obama 43%), Male 45% (HRC 34.7%, Obama 49.6%), White 62% (HRC 50.8%, Obama 33.4%), AA 32% (HRC 9.6%, Obama 72.5%).

____________________

boskop:

PAUL:

does this mean that there are more AA who are undecided in NC? It seems rather large and unusual at this juncture. please explain.

____________________

Paul:

In reply to boskop:

As to AA in NC, based on May 1-2 rolling poll: 14.4% of AA were not sure, and 3.5% of AA said someone else.

As to AA in IN, based on May 1-2 rolling poll:
8.8% of AA were not sure, and no AA said someone else.

One day before in NC, based on April 30 - May 1 rolling poll: 12.9% of AA were unsure, and 5.2% of AA said someone else.

In IN based on April 30 - May 1 rolling poll, 8.7% of AA were not sure and no AA said someone else.

For sake of comparison, three days before PA primary, AA represented 16%; HRC 4.8% of AA, Obama 89.8% of AA, not sure 5.4% of AA and someone else 0%. Last PA primary poll: AA represented 16%, HRC 11.7% of AA, Obama 80.6% of AA, not sure 5.6% of AA and other 2.2% of AA.

More comparison: last Ohio poll where 15% represented AA, HRC 8%, Obama 80.5%, not sure 10.5%, someone else 1%.

More comparison: last Texas poll where 20% represented AA, HRC 8%, Obama 78.1%, 9.3% not sure and 4.6 someone else.

____________________

Tybo:

"And you are assuming Maya Angelou or Senator Clinton will be a better role model than... ummm, the first African-American President of the United States of America."

Obama is just as much european-american(perhaps more because his mother family raised him)
He has no ties to the African American community.

TO vote for someone because of their race or gender is foolish, to vote for just skin color is the ultimate in stupidity.
Obama is counting on stupidity in many voters.

____________________

boskop:

PAUL!! wow, thanks
so let me see if i get this. based on the PA stats vis a vis NC, there is a rather large possibility that the unusually large undecided AA vote will break for clinton, correct?

if the passion for a fellow AA hasn't sealed the deal by now, it most likely wont?

if this is consistent with your data and i'm not totally reading this wrong, then it very well might be a game changer. i have found far more closet black women clinton people than the media tends to pick up. of course it's not numerically pure more of a street corner buzz kind of thing.

any way, many thanks for that terrific break down. it was fun.

____________________

Paul:

Given Obama's lead in NC, statistically speaking, the 14.4% not sure of AA should not make a significant difference. To put this into perspective, AA represents 32% of poll in NC. The 14.4% of AA not sure therefore only represents 4.6% of overall poll. Even if the AA not sure splits 60-40 in favor of HRC, the difference only amounts to 0.9% gain for HRC.

As to the prior post from Tybo, whether we agree or not, women are voting for the female, especially older women, and AA's are voting for Obama. Are they voting because of gender or race, or because the candidate's views match their views? How is it possible to separate gender, race or age from a candidate's views?

____________________

boskop:

PAUL:
this is the first time it just might not be possible to separate yourself from the candidate's profile given the variety of offerings.

that's why a minor change away from these kinds of tendencies is so interesting.

i certainly think obama will win NC. i do not think he'll coast. i think your stats will break other than you suggest. wishful thinking? maybe, but i have come to think that the race itself is the better part of this election cycle including whoever becomes president.

that's almost the anti climax. quite a derby i'd say, maybe a filly will photo finish today as well?

____________________

Tybo:

"As to the prior post from Tybo, whether we agree or not, women are voting for the female, especially older women, and AA's are voting for Obama"

there's a huge difference between 92% support and 65% support.

____________________

carl29:

I used common sense. After so many contest we already know who votes for who. I don't think the pattern will be broken any time soon. AA will keep on voting overwhelmingly for Obama and women, especially older, white women, for Hillary. Whites without college education for Hillary and whites with college education for Obama, especially educated, white men. People with good incomes for Obama, people with low income for Hillary. Old people for Hillary, young people for Obama. To predict who is going to win is pretty simple, just look at the demographics and there will be the answer.
NC: double-digit victory for Obama,around 15%, thanks to AA's and around 35% white support.

IN: Hillary's victory around the same as in PA, thanks to white women and white people without college education in general. Now, I think the margin will depend on whether Obama wins white men, which I doubt, and for how much he does lose them. I think that the swing demographic are white men.

____________________

RS:

@boskop:

Ummm... don't know where you got the idea that I have antipathy toward Senator Clinton's supporters. I only have antipathy towards illogical assertions.

FYI, Senator Obama never said he is against all wars - just dumb wars. So if he has to wage war - it will be a wise, considered decision.

____________________

Pollbuster:

A lot of people are fond of pointing out how Zogby nailed it in Pa. However the plain fact is that Zogby didn't nail anything, but rather he was trying to have it both ways. In his 2 day tracking survey, he had Clinton up 15 on the Sunday before the election, but on Monday, he had Clinton up only by 5--hence your 10 point spread. Of course if it turned out much closer, Zogby was prepared to say how he noted a big move towards Obama near the end. The plain fact is he was all over the board in Pa. because even on the Saturday before Clinton's big +15 Sunday, he had Obama at +2. It is difficult to give the Zogby poll any credence.

____________________

kingsbridge77:

Let's not forget that John Zogby's brother and senior analyst of his polling firm is campaigning for Obama by wrigint op-eds asking Hillary to withdraw from the race.

____________________

kingsbridge77:

By the way, CA was not the only state Zogby blew. He blew Ohio as well, and always in favor of Obama.

____________________



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