With anti-incumbent sentiment brewing before the midterms, Kristen Soltis is asking questions about Republicans, Democrats and Tea Party supporters.
Kristen Soltis is the director of policy research for The Winston Group, a Washington, D.C.-based polling and survey firm. Her research primarily focuses on election trends, young voters and political parties.
Politics: The Winston Group’s Tea Party poll took a long time and a lot of resources, is that what it takes to get an accurate picture of the electorate right now?
Soltis: In order to get an accurate view of the electorate you’ve got to make sure that you are putting in the time and resources to do it right. A survey is not just a commodity, not something that you can buy at a store and say, “Well I would like $20,000 worth of survey, so see what you can do for me.” If you have a question you want answered, it’s important to take the time and do the due diligence to make sure you get it right before you put it out there. In the case of our Tea Party poll, about 17 percent in the three surveys combined said they were Tea Party folks. Well, 170 out of 1000 people is not a size that’s insignificant. But to get a more complete picture, we wanted to see over three months if you push that all together, what does it look like then.
Politics: “Tea Party” is a bit of a loaded term on all sides of the debate. Does that make it hard to get an accurate reading?
Soltis: It depends on the question that you ask. Some pollsters ask “Do you support the Tea Party?” I believe there was a question some months ago, “Do you sympathize with the Tea Party?” We tried to ask a question that was pretty strong, so it was either yes or no, not how you feel about it. We wanted to get the most definite group.
Politics: Tea Party supporters tend to be a little bit older, and one of your specialties happens to be younger voters. At the end of the 2008 cycle, young voters were the hot commodity, has that flipped for 2010?
Soltis: I think that it’s not so much about whether younger or older voters are important. In midterm elections, the voter base tends to be just a little bit older. But again, for Republicans, if we ignore young voters this cycle entirely, we’re going to have real problems bringing them back into the fold in 2012 when their numbers go back up again. And I think Republicans have been given a real reprieve—it looked bad after the 2008 election. It really looked like young voters had just abandoned the Republican Party, and it looked like it would be nearly impossible to bring them back. Now the mediocre performance of the Obama administration in its first years has reopened the opportunity for Republicans. There was some research that came out from the Harvard Institute of Politics—they do a survey that I think is just fantastic—they are showing some areas of opportunities for Republicans. The fact that young voters are very concerned about the economy is one. Republicans need to explain what they want to do to fix the economy to bring young voters back into the fold.
Politics: The Tea Party poll found a surprisingly high percentage of Democrats as well as Republicans and independents in the mix. Does having everyone mad at Washington change the way you have to poll?
Soltis: Where folks who have said, “I don’t feel comfortable affiliating with one party or the other,” go is more interesting. A sort of little known fact—and I have it up here on my computer so I won’t forget it: In 1994, 34 percent of the electorate was Republican, and they broke for Republican candidates 91 to 8. In 2006, 36 percent of the electorate was Republican and they broke for Republican candidates 91 to 8. It’s not about turning out your base as much as many consultants would have you believe. It’s really independents going from in favor of Republicans by 14 points to against them by 18 points. That’s where you see the big shift. Right now, independents are looking a lot more like Republicans when you look at the crosstabs of the poll—and that’s very good for us.
Politics: What can national parties do in an election when nobody wants to be seen as “from Washington”?
Soltis: I think what national parties can do is work on the policy side, on the ideas side. Here in Washington we have such a wealth of think tanks and people on the Hill who are real policy experts. Rather than sitting back and letting ourselves be branded as the “party of no,” Republicans should be taking this time to come up with policy solutions that resonate with voters, that energize independent voters and re-establish that trust with independent voters which we have lost over the last couple years.
Often you hear that folks inside the Beltway can be disconnected —an example of this was the healthcare debate. If you asked voters what their top issue was economy and jobs were clearly out in front, no questions asked. Yet folks in Washington, the Obama administration, said now is the time to tackle healthcare. And poll after poll kept showing that was the wrong topic—not necessarily that voters were for or against the policy, they ended up not supporting it, but that they were on the wrong topic for so long while unemployment was staying so high.
Politics: There are so many polls that come out so often now—whether they’re automated or dialed calls— and there are so many questions about the validity of polling. What do you look for when you try to measure a poll’s validity?
Soltis: We always like to err on the side of caution, and bigger is always better in this case. Oftentimes campaigns have budget concerns and selling them on that extra 100 people in the sample is often hard, but it’s critical. If you’re looking at a primary and general situation, you need a big enough sample of people who are Republican primary voters if you want to be able to make judgments about both where the primary and potential general match-ups stand. Particularly with the first survey we do for a campaign, we tend to push for a sample size that would be a little bigger than what they might want for their campaign’s budget. For a congressional district, we’re looking at about 400. A lot of people go with 300 because that will fit their budget, but we prefer 400.
Politics: Do you scoff at automated polls?
Soltis: We do live telephone interviews, but right now no methodology is perfect. No matter how much pollster may want to say, “What I do is the best, what I do is the gold standard,” there really is no perfect way to do it right now with the increase in cell phones or Internet versus live interviews. As someone who is younger in the industry, I’m very concerned about what my job will look like ten years from now. We probably won’t be doing heavily land line, live-interview telephone surveys.
Politics: If you had to predict, what would the next model of polling look like?
Soltis: When survey research began, there was a bias in surveys because not everyone had telephones, and now it is taken for granted that everyone has telephones. Now Internet and cell phones are sort of the new era. The Internet—it isn’t going anywhere, and eventually it will be as much of a given in an American household as the landline telephone was 30 years ago. The questions of how can you effectively sample, how can you cut through the clutter of spam and pop-ups—these are all important challenges that the industry has to face—but the Internet is definitely going to be the wave of the future. The folks who can figure it out first are going to be the ones who come out on top.
Politics: Have you tried out some of the Internet models?
Soltis: We ran an experiment. We used live telephone interviewing and simultaneously ran online questionnaires with an online panel. With weighting—minor weighting—we were able to get the data to line back up. It was something new for our firm to try, and we’re still working on it internally as to how we want to use that. The capabilities are so fantastic. Think about it, instead of doing a focus group sitting around a conference table with 10 potential voters and clicking through a DVD, you can get a much larger sample size—something that sort of transcends the boundary between qualitative and quantitative research where you’ve got a larger sample size that can make you more confident in what you’ve found. And the Internet should theoretically be able to let people do research cheaper as well. I think one of the things I have enjoyed most about working at the Winston Group is that we look at polling differently than we did five years ago. We’re trying to figure out ways to look at things like game theory—where if I do strategy X and my opponent goes with strategy Y, what’s the payoff in terms of voters? If 100,000 people listen to message X and Y, who comes out on top? I think you can use polling to begin getting at some of the very quantitative goals for a campaign to reach and understanding how your goals and strategies might interact with what your opponent might do. I really see us trying to integrate more of that thinking into our polling. It’s really going to go beyond just message testing and ballot testing.
Politics: Can you ever have too much information or too much data?
Soltis: A lot of pollsters will give a very data-rich PowerPoint, where at the end of it, you’ve been given 100 pieces of very important data but 100 important pieces of data winds up being a blur. Here at the Winston Group, our presentations look a little bit odd. Rather than big, crazy slides loaded down with data, we’ll just have a picture up because it’s got to be part of a narrative. Too much information can wind up being no information at all if the themes and narratives that link it all together haven’t been identified. You can have a whole book of cross tabs, but if you can’t pick out the five or six numbers that mean the most to your campaign, it’s just a heavy binder. If my client is only going to walk away from his remembering two or three things, what are the three points—supported by the data—that they need to remember to achieve success.
Politics: Do you feel like you’re competing with public polling? Is it hard to keep track of the message with so many other polls out there?
Soltis: No, because I think what we provide is different. For instance, this Tea Party memo is one of the first times that the Winston Group has really embarked on a public study that we are releasing and really driving. Prior to that, it’s been talking to reporters and influential people here in DC, pushing our message on how the party can win through those channels. As far as completion for eyeballs, people are just hungry for data. They just want to know what’s going on, and anyone who can provide data that they believe is credible is going to be able to attract those eyeballs. What we’re able to provide is different from what your average public pollster wants. What the public pollster is looking for is the eyeballs, but we want the eyeballs plus the answers that can create strategy. Rather than something that just creates an interesting headline, we want an interesting piece of knowledge that becomes a message.