Scott Rasmussen Explains The Latest Polling
HH: It’s my favorite pollster, Scott Rasmussen, of Rasmussen Reports. It’s so silly now. Every 6:30 in the morning on the West Coast, I clue in to Rasmussenreports.com to find out what Scott is telling us about the tracking poll. Scott, today it was a gift-wrapped surprise for Romney supporters like me. He’s up four in your tracking poll. Are you surprised by a two point jump overnight?
SR: I was very surprised, and especially because this race has been so close for so long. If you took out the convention bounces, neither candidate has had a four point lead in our tracking poll for months. So this is a really good day for Governor Romney in our poll. Having said that, I know you want to treat it like a present, but when something jumps out at you like that in a poll, you want to wait a few days to see if it’s real, or if it’s a just a little bit of statistical noise.
HH: Now Scott, there’s a lot of statistical noise out there. And there are a lot of pollsters out there, I know you don’t slam your competitors or your colleagues in the business, but John Podhoretz and I, and a number of other people, we’re just…besides you and Gallup, I don’t trust any of these people. They don’t do it often enough, and they’re not transparent like you are. Is this cycle helping or hurting the reputation of pollsters?
SR: Well, the reputation of pollsters always varies depending on whether your team wins or loses. People who think there’s something wrong want to blame the pollsters for the outcome. I even had somebody tell me that after Scott Brown won, a Democrat said it was my fault, because we made a mistake, and we polled showing he had a chance to win, and that changed the outcome of the race. So I think overall, there’s a lot of confusion about the polls, and most of the reason that there’s a problem is because the reporting of the polls is not very good. If you look at the outliers, and you think oh, my gosh, here’s where the race went from yesterday to today, you see something that makes no sense. If you look at the common ground in all the polls, every poll by every firm, no matter how they approach this, they’ve shown Governor Romney doing better since his first debate. They show improvement in the swing states. And they continue to show the President in the high 40s in terms of his support. And that tells me that you count on all those things, the President is struggling right now, but not totally out of it.
HH: Now I try and go about this professionally and like a lawyer, Scott. And the two questions which some of your colleagues cannot answer is the cell phone weighting and the refusal to weight samples. So let’s take those in turn. What is the cell phone dilemma, and how are you handling it in this cycle?
SR: The cell phone dilemma is much bigger than anybody really wants to acknowledge. And the reason it’s much bigger, it goes back to why telephone polling used to work. You, like me, are old enough to remember when some people would say sssh, it’s long distance. You know, we took those calls seriously. Younger people don’t even know you can use a cell phone to talk. They Tweet, they text, they do everything else. So calling them on a cell phone is not really a normal conversation to engage them in. We need to find ways to catch people where they are having communications. Ultimately, that means we have to figure out how to tap into Facebook and Twitter, and other social media. Right now, at Rasmussen Reports, for people who have abandoned land lines, we are reaching them through an online tool. We have a sample that is provided by a third party that gives us an absolute airtight way to validate their demographics, to make sure if somebody tells us they’re a 22 year old female, they really are a 22 year old female. It’s imperfect, but it is a good approximation. And what we’re finding is when you include that cell phone sample in the overall numbers, it shifts the numbers, it shifts the polling about a couple of points in the Democrats’ direction.
HH: Now I also have been trying to get…there are monks of different pollings from different orders, but I’ve brought them all on, and I’ve asked them about sample sizes, which is it’s not an attempt to indict them. I just don’t understand how you can have a nine or a ten or an eleven point Democratic sample size. What’s your theory? How do you approach sample sizes when it comes to these very close races?
SR: We do adjust our polls for our model on what the party identification should be. We are able to do that because we conduct so many polls through the course of the year, we measure consistently how much, because people do change their party identification subtly over time. Some people go from being a soft Republican to an independent, or an independent to maybe a soft Democrat. So we track these movements, and we then apply them to our national model. And when we go into the states, we do the same things. Typically, the impact of that is to sort of smooth out the bounciness of it. But it does avoid us having a situation where a state that was even in the partisan split four years ago all of a sudden has a fifteen point edge for one party or the other.
HH: All right, so I think that is absolutely essential. It’s persuasive to me, and some people don’t do it. So that’s part of my critique. Now let me ask you about enthusiasm, and how that, A) what are you seeing on the enthusiasm scale as between Romney and Obama, or Obama and Romney? And how does that impact your approach to the polls?
SR: The biggest single indicator of whether somebody’s likely to show up and vote is how engaged they are in the campaign, how closely they’re following it. And think of it like a sports team. You know, if your team is doing well, you go to the sports page more often, because you just want to read everything about it. When your team is kind of letting you down, you don’t have that enthusiasm. Right now, in this election, every indicator that we have shows Republicans and Republican-leaning constituencies are more likely to turn out and vote than comparable Democratic constituencies. And in a couple of areas, the youth vote is still going to go for Barack Obama, but the enthusiasm is down. The Latino vote is still going to go for Barack Obama, but the enthusiasm is down. On the other side, seniors in particular much more engaged this election cycle than four years ago. And by the way, we’re aging as a society. There’s more seniors. It would not surprise me at all to see turnout among seniors up a little bit, and that’s very helpful to Governor Romney. Hugh, when you put it all together, I just want to give you a sense of scale on all this. In 2004, 37% of voters were Republican, 37% were Democrats. Dead even. In 2008, it was 39% Democrats, 32% Republicans, a seven point Democratic advantage. What we’re estimating right now, what we’re seeing in our polling is the Democrats are likely to have maybe a two point, perhaps a three point edge, somewhere along the line of 36% or 37% Republican, 39% Democrats. And if that’s the case, Mitt Romney has to do fairly well among the independents to win. Right now, he’s doing that.
HH: You know, I’m a Cleveland Browns season ticket holder, Scott.
SR: I’m sorry.
HH: And as the season goes on, I know, it gets harder and harder. You leave two tickets on the windshield, you come back and there are four. It’s hard to give them away. Is that what is happening right now, we’ve got a minute left, on the Democratic side? Are you seeing them draining away their passion? Or is it stable?
SR: Right now, it has stabilized. After the second debate, the President did manage to stop the bleeding. But he did not manage to turn it around, at least so far. You know, we don’t know what the final debate will have done. My gut feeling is it probably doesn’t change things very much.
HH: Scott Rasmussen of Rasmussen Reports, thanks for joining us. Always the guy who is willing to answer any question directly, on point, and concisely. I’ll let people compare and contrast that with the other interviews I’ve done with pollsters throughout this campaign.
End of interview.