printable banner

U.S. Department of State - Great Seal

U.S. Department of State

Diplomacy in Action

Latest National Polls and the Upcoming Presidential Debate

Peter Brown, Assistant Director of the Quinnipiac University Polling Institute
Washington, DC
October 2, 2012

12:30 P.M. EDT


MODERATOR: All right. Good afternoon, and welcome to the Foreign Press Center. We’re very happy to have with us today Peter Brown, the assistant director of the Quinnipiac University Polling Institute. Peter’s here today with us to share the results of Quinnipiac’s latest national poll and give a brief preview of tomorrow’s presidential debate.

So with that, I’ll just open it up with Peter.

MR. BROWN: Good afternoon. How are you today? I’ll do a very brief summary of the poll, and then we can take questions. Latest Quinnipiac poll finished, which finished Sunday night, showed the President with a four-point lead, 49 to 45 percent among likely voters nationally. The President’s ahead because women love him. He’s got an 18-point margin among women. Romney has a 10-point margin among men. That’s not – the notion of the gender gap is well entrenched in American politics, but the size of the gap this year is unusually large.

The President is getting – the President is winning the African American vote 94 to 4, losing the white vote roughly by 11, I think it’s 53-42. Nothing surprising there; that’s kind of been the case for a while. The 42 percent of the white vote that the President is getting is roughly what he got four years ago. Exit polls in 2008 showed he got 43 percent of the white vote. If the President can stay in the 40s among white voters, he’s very, very likely to get reelected. If it starts dropping into the 30s – which it was in some polls earlier this year when he and Romney were close and – or Romney had a slight lead – then that could be a problem for him. Whites make up 72, 73 percent of the electorate roughly.

One of the interesting findings is that voters say that they think one-party rule in Washington would be better for solving problems. Historically, voters have been reticent to want one party to run the whole show. They have always viewed such things suspiciously. They like one party checking the other one. But that probably reflects frustration with the inability of the federal government to deal with problems that the voters think should be dealt with.

On the debates, 9 in 10 voters say they’ll watch it. Probably won’t be that high, but that’s what they say. Roughly 1 in 10 say it could change their minds. If they all do, it could be important. And by 2 to 1, they expect the President to win the debates, which is probably the best piece of news in this poll for Mr. Romney, since it gives him a low threshold that if he were to exceed by a lot might help him and might pick up some votes.


MODERATOR: And I just want to remind everybody to state your name and media organization before opening with your question.

Right here.

MR. BROWN: Yeah.

QUESTION: Hey. This is Nico Pandi from JiJi Press. Thanks for coming in. I was just wondering if you could tell us, have you noticed any kind of narrowing in any of these groups between, like, the white vote and how much support Obama or Romney have had or --

MR. BROWN: Narrowing from when?

QUESTION: Since Romney became the official nominee.

MR. BROWN: Oh, back in the early spring?


MR. BROWN: Since the convention or back when he clinched the nomination?

QUESTION: When he clinched the nomination.

MR. BROWN: Yeah. I mean, Romney was running a little bit better against the President. But again, we weren’t doing likely voters then, so that’s always a problem, because we were doing registered voters. But again, the President’s inched up among whites, which is very important to his reelection. He’s got to be in the 40s. If he’s in the 30s, he’s got a problem, and he was in the 30s among registered voters for quite awhile this spring. So that’s a big thing.

Again, 94-2 among African Americans. I don’t know where the two are. He’ll get – that leaves – that adds to 96. It’s a good bet that he’ll get almost all of the remaining four. I mean, the size of the African American vote and the – and its turnout numbers four years ago were important to Obama’s win. In the end, the race wasn’t that close, but the expectation was going to be – was driven by the fact that he expected and got an extraordinarily large African American turnout.

Normally, black voters go 90-10 for a Democrat. He’s going 97, 98. That makes a difference in key states – places like Virginia, which has a 15 percent, roughly, black population. Ohio has a little bit less than that. Florida, a little – is about 12. Those are states that are key swing states where large African American turnouts are a big deal in terms of helping him carry the state.

Yes, ma’am.

QUESTION: Hi. I’m Camille El Hassani from --

MODERATOR: Please wait for the mike.

QUESTION: Sorry. Camille El Hassani from Al Jazeera English. A couple of questions. I was wondering about turnout. I haven’t been through the whole study here, but do you have a sense of – is there a sense of any sort of voter apathy amongst those people who just aren’t going to turn out?

MR. BROWN: We don’t seek to measure turnout. It’s a very difficult thing to do, and it’s really not what pollsters do. But here – there is a question you might want to look at – hold on – which is the enthusiasm question, and I’ll find it for you in a second, hopefully.

Question 12: Compared to past presidential elections, how would you describe your level of enthusiasm about voting in this year’s presidential election? Forty-four percent say more enthusiastic; 14 percent say less.

That’s a little bit of a deceptive figure, because there is some split by party. Fifty-five percent of Republicans say they’re more enthusiastic. Not surprising; the out party is always enthusiastic about trying to get the incumbent out. Democrats are 41-10 more enthusiastic. That’s better than it has been. Democratic excitement has increased over the last few months. And independents are roughly this – are 4 – 2 to 1 they are more excited. So people say they’re more excited. It will be hard to see larger turnout than four years ago, candidly. But we’ll see.


QUESTION: Hi. Tao Zhang from China, Caixin Media. I mean, based on the historic experience, how largely can the debate influence the polls? After the debate, I mean.

MR. BROWN: They can be a big deal. If Gerald Ford hadn’t freed Poland in 1976, he probably would have gotten reelected. He made a big boo-boo. Everybody familiar with that? Clearly, the debates mattered there. There was a general sense in the primary debates this year, for instance, that the governor of Texas wasn’t up to the job because of his debate performance. What people tend to remember is goofs by candidates. But a candidate who is – needs to pick up a few points, if he were to dominate the debate and – it might help him.

Again, 90 percent of voters say they’re going to watch. Ten – I think it’s 11 percent said that it could change their minds. I don’t think it’ll be 90 percent, and I don’t think – but a lot of voters are going to be there.

What is true is it’s clearly Romney’s best opportunity in the remaining time. He’s got three debates, going to have 50 million people watching each of them. It’s a pretty big audience. Whether it changes things or not, we’ll see. But it is an opportunity and probably his best one.

Yes, ma’am. I couldn’t see you.

QUESTION: Garcia, I’m with Notimex. Do you believe that the first debate is definitive for Mr. Romney, or –

MR. BROWN: I’m sorry, I’m having trouble understanding you.

QUESTION: Do you believe that the first debate is definitive for Mr. Romney or – definitive --

QUESTION: Decisive?

QUESTION: -- decisive, very decisive?

MR. BROWN: It could --

QUESTION: Depending in a good performance or a bad performance, it could be that’s it for him?

MR. BROWN: Well, look, he’s behind, and this is the best opportunity to change that. Whether it will or not, that’s the $60 million question. We’ll see starting Wednesday night. But it’s an opportunity to be exposed to a huge number of voters when some voters who are not all that interested in politics finally begin to pay attention. So I’m not trying to repeat myself, but whether it works – whether it allows him to change the status quo or not, it’s an opportunity to change the status quo. We’ll see whether he takes – is able to take advantage of it.


QUESTION: Hello, this is Ana Sanda (ph) from the Romanian Public Radio. I am wondering what could Mr. Romney do in order to change things in the --

MR. BROWN: If the vote – if voters perceive him as more competent, as offering an argument they hadn’t heard before, or perhaps if he shows himself to be something that they thought he was not – again, 50 million people is a lot of eyeballs, and it’s his opportunity to change the status quo. If the election were today, all the polls say he’d lose. But it’s not today. It’s five weeks from today. And this is his best – this and the other two debates are his best opportunities to change the status quo. Tie goes to Mr. Obama. Mr. Obama would be thrilled if these debates are judged as a tie because he’s ahead. Romney has to win the debates.


QUESTION: Thank you. Xiahong with China Radio International. And my question is: What do you estimate will be the issues that people – that the voters are likely to hear on the first debate?

MR. BROWN: Well, this one’s supposed to be on domestic affairs, which means it’ll be about the economy, one assumes.

QUESTION: What about – beyond the economy, what are the other topics, like tax cuts or –

MR. BROWN: Well, tax cuts, spending cuts, how you deal with the debt, taxes, healthcare probably, which is an economic issue to a lot of people – those are all issues that are likely to be discussed. Again, we who do politics for a living in journalism are very conversant with what the various positions are, but the housewife in Omaha might not be, or the guy who doesn’t tune into politics until it’s just before the election. Their level of information may not be what many of us have as a level of information, and this is an opportunity for them to find out things that they may not know about the candidates.

MODERATOR: I’m going after a New York – can we take a question from New York?

MR. BROWN: Sure.

QUESTION: Good morning. Marta Torres. I don’t – can you hear me? No, he cannot hear me, Jon. He cannot hear. Yes?

STAFF: Yeah, they can hear you.

QUESTION: Okay. Good morning. Marta Torres from La Razon newspaper from Spain. I would like to know why still there are – we are like, as you said, five weeks to go for the election and still Romney has to somehow change the game. Why – what’s wrong with his campaign that he hasn’t been able to change the game for the last months?

MR. BROWN: Mr. Romney has not had a particularly good month. The American people have decided, apparently, that the television show that the President and the Democratic Party put on in Charlotte was better than the television show that Mr. Romney and the Republicans put on in Tampa. In other words, they won the convention phase of the campaign. The 47 percent furor over the last two weeks certainly hasn’t helped Mr. Romney. It’s gotten him off message.

But what you’re seeing is a presidential campaign. Issues come, issues go. The one thing that Mr. Romney has not been able to do is shake Mr. Obama’s lead among who voters like better and who they see as being more interested in and in tune with their values and lives.

QUESTION: Hi, my name is Andrei Sitov. I am with Tass, the Russian News Agency. Some of the Republican partisans, like Governor Christie for instance, say that the day after the debate we’ll see a completely new situation, it will be a game changer, all of that. Historically, have you ever – doing polls, have you ever seen such a sea change after one debate?

MR. BROWN: Gerald Ford giving – freeing Poland seemed to make a huge difference.

QUESTION: Excuse me. But from what I heard about that debate, many people did not even realize he made that mistake.

MR. BROWN: Yeah, but they sure found out about it quickly.

QUESTION: Until – yes, until after the journalists in the next week or so explained to them what happened. And I think he was even given a chance by the presenter, by the host, to extricate himself from that situation, but he never used it. His mistake, basically, was not making the mistake. His mistake was being stubborn about it.

So anyways, other than that, you are not – you cannot think of a single instance where –

MR. BROWN: Well, I mean --

QUESTION: -- thank you.

MR. BROWN: There are people who think that Al Gore threw away the presidency with his debate performance in 2000. Obviously, that was – you can’t get closer than that election, so anything that shifted 500 votes determined – in Florida determined the election. That debate, obviously, was – mattered a lot. 2004 to 2008, no. ’96, no. ’92, not really. But again, you had a third party in ’92 and ’96, and that really messes things up. ’88, Dukakis being unwilling to want the death penalty for a killer – someone who killed his wife probably didn’t help him, but he was behind before that, but that certainly cemented it.

’84, Mondale thought he had scored big in the first debate when he – when Reagan appeared to some to be not quite with it, but Reagan came back in the next debate and ended all speculation about that. In 1980, the belief was – the consensus was that Reagan won the debate with Carter and that obviously made a big difference. Carter was ahead in the polls up until late September. So the debates played a big role in ’80. And then we’ve got ’76. I don’t even remember if there were debates in ’72; I was pretty young then.

QUESTION: Not in late September.

MODERATOR: All right. Can we go to New York?

MR. BROWN: Hi there.

QUESTION: Yeah, hi. It’s Maria Ramirez from the newspaper El Mundo.

A couple of questions. From what you saw in the – in terms of style of the two candidates from previous debates, what can we expect tomorrow tonight? And secondly, are there any evidence that the debates are more watched in swing states? Thank you.

MR. BROWN: I don’t know. Probably not. I mean, I don’t – I just don’t know about whether they’re more watched in swing states. They’ll be watched by a lot of people, both in swing states and non-swing states. Obviously, the swing states are more important, but I have no way of knowing whether the comparative relative viewership there will be greater on a per capita basis than there will be in Texas, for instance, which is not a swing state.

Your first question? First part of the question?

QUESTION: Yeah. In terms of style of the candidates from what you saw in previous debates, in Romney in the primaries or Obama four years ago, what can we expect? What sort of debate are we going to have tomorrow?

MR. BROWN: Hard to know. Candidly, Romney was debating a bunch of other people who were not the President of the United States. And in a primary, the debates tend to focus more on narrower issues. Obama seemed to do – when – you could argue that Obama won the debates against John McCain four years ago, but John McCain’s not Mitt Romney. It’s been four years. I mean, I’m probably not the right person to critique their styles. That – I’m good with numbers. I’m not good with stylistic stuff. Not my job.

Yes, ma’am.

QUESTION: Hi. On the issue of undecideds, you’ve got here that about 5 percent might change their mind. Is that a typical number for this time in an election?

MR. BROWN: It’s probably a little lower than normal, but normal’s tough to – I mean, one of the things about polling is it deals in numbers, and one of the reasons why people look at polls, in some polls and say the N isn’t big enough, another you don’t have enough cases to have a statistical thing – we don’t have enough cases on presidential elections to have a statistically valid way of saying, “Yeah, there’s more this time than last.”

But it’s – look, you have an incumbent. Elections with an incumbent tend to be about the incumbent. And most Americans have a pretty good view of what they think about Barack Obama; less so Mitt Romney. And people either like him or not, and that probably is one reason why you have such a low number of people who say they’re undecided.

On the other hand, if something dramatic happens at the debate to change their view of Romney and Obama, that would theoretically be the kind of event that might shake things up.

Yes, sir.

QUESTION: Hello, Mr. Brown. I’m Mark Yates from the Australian Broadcasting Corporation. I’ve got a couple of questions about this suggestion that pollsters like yourself are oversampling Democrats and under-sampling Republicans. My first question, really, is what is your response to that criticism? And then could you talk us through a little bit about how you get these accurate samples, how –

MR. BROWN: Sure. Let me --

QUESTION: -- you reflect the balance? Okay.

MR. BROWN: Let me give you polling 101 very simplified. Okay.

We and virtually every other poll in the country buys phone numbers – or excuse me, phone exchanges from a vendor. And what they give us is – if we say we’re going to – if we want to sample for Ohio, we buy a sample of phone – a statewide sample of phone exchanges from around the state, which means that we will get electronically delivered a list of phone exchanges which would be – for instance, let’s say if it’s in Cleveland, 216, which is the Cleveland area code, and then the first three digits, and then our computers will randomly generate the last four digits of the phone number.

And we continue with the – keep calling back that phone number if it doesn’t answer or if somebody’s not home. Now, a couple – the most important thing to a sample is that it is random; in other words, if the total sample is an accurate representation of the state that you’re polling. Not just in Cleveland, but the whole state. Okay.

So here’s what we do. Our callers have a script they read. And here is what it kind of sounds like: “Hi, I’m Joe Blow, I’m calling from the Quinnipiac University poll in Hamden, Connecticut. We’re interested in your views on the issues of the day.” We’re currently in a deal with the New York Times and CBS, so that script has been modified a tad to include CBS and the New York Times. And then we’d say, “Can I talk to the next person who has a birthday in your household?” Why do we think we do that? Anybody got an idea?

Here’s why: Because women answer the telephone more than men do, and we want a random sample of people, so the best way to do that is – we found what we call the birthday method. And if the person who has the next birthday in the household isn’t there, we say thank you and we leave a message and say, “We’ll call back later. When do you think they would be there?”

Okay. So that’s how we get our sample. Then we try to find out – then we ask people if they’re registered voters. And earlier in the year, all we’re interested in is registered voters. We only do registered voter polls. We switch over to likely voter polls generally four or five months before the election. We’re now in the likely voter scenario.

So we have this voter who has said that – they have the next birthday, and they are in this household with the phone number and we have called. And then we ask if they’re registered to vote. If they’re not registered to vote, we thank them and we go our merry way, because if they’re not voters, we don’t care, to be blunt about it. There are some organizations out there that survey adults, but we don’t.

Now at this time of the year, we’re interested in likely voters. So we ask these people who are registered voters a series of questions about their past voting behavior and inclination to vote. That, then, becomes our likely voter sample. If someone passes the screen, we include them as a likely voter. If they don’t, we don’t, because we don’t think they’re likely to show up. And if they’re not going to show up, we don’t care. Okay.

So why are we having more people who say they’re Democrats than Republicans in the sample? And the answer is that voter ID – in other words, when we call, we ask voters whether they consider themselves a Republican, a Democrat, an Independent, or a member of some other party. That’s a changeable figure.

I’m happy to report that two years ago, I had the governor of Ohio personally attack my poll because he said we didn’t have enough Democrats in our samples. Well, how many people were here in 2010? Raise your hand. What’s different between 2010 and 2012? 2010 was a Republican year. 2012 is less so. So that – his complaint is we didn’t have enough people who said they were Democrats. They didn’t say they were Democrats because they didn’t feel like Democrats. People’s voter identification changes over time. It goes like this: back and forth, back and forth. Okay. Not everybody, but a number of people.

So we do not weight our samples for party identification. In other words, we don’t say, “The sample should be 32 percent Democrat, 31 percent Republican, 33 percent Independent.” It is what it is, with one exception. And we don’t weight party ID because there is no standard to weight it to, as you – as I’ve just tried to explain. But we do weight samples for immutable characteristics: age, sex, race – I’m probably forgetting – education – because there are standards we can check our data against. And that’s the U.S. Census Bureau data for the particular state that we’re polling at the time. So that if, for instance, the state is 14 percent African American and we only have 9 percent African Americans in our sample, we weight the sample to reflect a 14 percent black sample.

Now, sometimes, when you weight for age, sex and gender, and race – especially race – it changes the horse race numbers. If, for instance, we have to weight up our sample to make it more African American, that is going to increase the Democratic ID of the sample, and it’s going to increase the Obama numbers in that sample. That’s because African Americans are a 95-5 constituency. They’re 95 percent Democratic roughly. Does everybody follow me?

That’s what we do, and that’s what most good pollsters do. There are some pollsters who try to make believe they know what the sample – what the electorate is going to look like and weight it. We don’t do that. Most of the very good polling organizations – the television networks, the Times, Wall Street Journal, us, Pew, Gallup – we do not weight by party. And that is the issue that has led to the controversy that you alluded to. And I’m sure that’s much more than you wanted to know.


QUESTION: I just had a brief follow-up question, then. So what do you say is behind this criticism that you’re not sampling? Is it just cynical politics?

MR. BROWN: The same thing that was behind the Democratic criticism in 2010. They don’t like the numbers. It’s – there’s no – the Democrats were just as vociferous two years ago. We – by the way, as I said, the governor of Ohio made a special case of going after our numbers. He’s the former governor of Ohio.

MODERATOR: Can we go again to New York?

MR. BROWN: Sure.

QUESTION: Hello. My name is Javier Boreli from Tiempo, Argentina, Buenos Aires. I wanted to ask you, regarding the polls in the swing states, what do you think the Republican Party should emphasize in the following debates? And considering that, in which debate we can look carefully about their statements in order to reach these votes they need?

MR. BROWN: I’m really good about telling you what the numbers are. I don’t give advice to candidates. Our polls show that voters are most interested in the economy. But they’ve been most interested in the economy for a long time. I mean, I don’t think the debate will be any different than the campaign has been. It’ll be about the economy, the economy, the economy, something else, the economy, the economy. I mean, I – it would be hard to imagine they’ll bring out anything spectacularly different at this point.

There’s an old expression in some of the swing states, which is: “You dance with the one that brung ya.” The economy was Romney’s calling card in the primaries; it was his calling card when he was ahead of Obama back in the early spring. I’m sure he’ll stay on the economy.

MODERATOR: Any other questions?

MR. BROWN: Yeah. There’s somebody coming up in New York.

MODERATOR: Ah, all right. All right, well, first let’s go here for someone in the room.

MR. BROWN: Yeah.

QUESTION: Yeah. Xiahong, China Radio International again. A lot of people are talking about this campaign saying that the election will be decided by those undecided voters right now. So what is the percentage of these people – these undecided voters around the population? And what influence do you think the debates will do on these people? Thank you.

MR. BROWN: Well, our polls, as one of your colleagues mentioned earlier, show only about a 5 percent undecided. Now I don’t know – I haven’t actually – didn’t look – I’m quoting you guys. I can look at the --

QUESTION: (Inaudible.)

MR. BROWN: Forty-nine, forty-five would leave you with six undecided. There are some people obviously in the 49 and the 45 that are soft, and obviously the Romney people hope to pick off some of the people who are soft Obama supporters and vice versa. But a solid percentage of those who say they’re undecided won’t vote. That’s what history tells us. They just don’t do it. And obviously, who goes to the polls, if it’s a close election, will matter enormously. If it’s not a close election, it’ll matter less.

MODERATOR: Can we go to New York?

QUESTION: Hi. I’m Nadia from South Africa. I wanted to find out how soon will we start to see the influence these debates – this debate is going to have on people?

MR. BROWN: I am sure that there will be organizations that will do polling after the debates. Everybody has their own way of doing things. I would caution you to – snap polls are not always reliable. One and two-day samples tend to be less reliable than longer samples, which give voters an – remember what I said about if the birthday person isn’t home, we call back? Well, good polls need time to call back those people to get them, because they’re part of the random sample. If you go – if you were to go to a new group of people, that would destroy the randomness of your sample.

So we generally go five or six nights. Some people do it quicker. It’s a matter of what you think is best. We like longer. Now obviously, that’s difficult with a continuing series of polls – a continuing series of debates over the next three weeks. Some people don’t poll during debate times because they think it’s meaningless data. It goes like this, based on whether there’s a debate or not, et cetera.

Yes, ma’am.

QUESTION: How many more of these polls are you going to do before Election Day?

MR. BROWN: We don’t talk about our polling schedule.


MR. BROWN: We talk about a lot of things, but not our polling schedule. (Laughter.)

QUESTION: Can we expect any?

MR. BROWN: Let me put it this way. We poll 49 weeks a year. We shut down for Christmas. We shut down parts – we shut down for Thanksgiving. We shut down for a day here and a day there. It would be logical to assume that we’re not shutting down between now and November 6th. But as to where we’re polling and when, that’s – yes, we will be busy.

Yes, ma’am.

QUESTION: My name is Julia. I work for an Austrian daily newspaper, Kurier. I have a question about the voter turnouts.

MR. BROWN: About what?

QUESTION: The voter turnout, the turnout of voters at the election. How many voters are – have you polled how many people intend to go and vote in November? And have you broken those numbers down by different voter groups – women, African American --

MR. BROWN: Well, again, they’re not in here if they don’t – if we don’t think they’re going to vote. Again, this is a likely voter poll. We don’t include data from their preferences if we don’t think they’re going to show up. So that this poll of likely voters is – we think – obviously it’s not going to be 100 percent, but we – essentially, we’re saying these are the people who are going to vote, and this is a sample of them. And they all say they’re going to vote. Most of them we think will.

QUESTION: But how many? 60 percent, 70 percent?

MR. BROWN: Oh, of the likely voter sample? One would assume a very large number – a very, very large number. I’m not going to throw out a number, because it would be – but a very, very large number.

QUESTION: It’s just interesting to compare between – I was interested in comparing the figures of 2008 when you had an exceptionally large turnout. What is expected this year, the --

MR. BROWN: I’m not sure what – are you asking me whether – what we think – do we think turnout rates will be higher than they were four years ago?

QUESTION: Follow-up --

MR. BROWN: It would make sense that they would be a little lower. Not because of any numbers we have, just because of human nature and because President Obama is – and I don’t mean this in a negative way, necessarily – a polarizing figure. Most Americans either really like him or really don’t. But I can’t give you a number.

MODERATOR: Any final questions? Oh, way here in the back.

QUESTION: Thank you. My name is Ivica Puljic. I’m from Al Jazeera Balkans. Thank you for doing this, and my question is: Do you know any example that – I don’t know – some candidate for president was really like – looked like he’s winning and on the end he was a loser?

MR. BROWN: I’m sorry. I –

QUESTION: Do you have, like –

MODERATOR: Who’s winning and then lost the election.

MR. BROWN: Well, Jimmy Carter was leading going into October of – in 1980 and lost. There was just one debate that year. In ‘88 Bush, was ahead. Dukakis’s poor debate performance just widened the margin. ‘76, Ford was way behind and made up a lot of ground in the last few – month and especially the last few days, and lost very narrowly. Again, had he not freed Poland and had the media paint him as stupid for quite a while, he might have edged Carter out instead of the other way. It was a very close election.

QUESTION: Thank you.

MODERATOR: Okay. If there are no further questions, this briefing is concluded. Thank you, Peter, and we look forward to seeing you again sometime.

MR. BROWN: My pleasure.

# # #