More Than Numbers: How Qualitative Research Can Win Elections

This piece was first published on April 3, 2013.

You led a new initiative this election with the Obama campaign: online focus groups. How do you think this provided you a leg-up in the campaign?

We had something called the online community, which was an ability to check in with voters on a regular basis with regard to events that would pop up without notice. The online community was an opportunity to get qualitative feedback from swing voters across the country in the swing states that we chose, and allow them to interact with each other but also to give us a very quick response to stimuli. I think the most important positive of the online community is that it gave us a speed by which we could get feedback.

Traditionally in the world of public opinion, Joel Benenson puts a poll in the field, which is a quantitative telephone survey that usually has a three to four day turnaround. The difference with the online community is that it was always available, 24/7. So when there was a breaking event, if the President said something in a press conference and all the sudden it’s aflame on Twitter, you want to see how the regular voters takeaway. We have the ability to find that clip from the press conference, put it up in the online community, and then within 60 minutes people are responding, and we’re getting a lay-public view of reactions. So part of it was timing, that we could get immediate reaction to events that are unpredictable.

The other thing was that it allowed voters to interact with each other in a continuing dialogue in the way that an online chat would, except that these people don’t know each other. That allows us to look at the language being used as people talk about the issues that we want them to talk about. So instead of transcribing a focus group or an open-ended question on a poll, we open up the community and there is the dialogue. We can cut and paste their own language, and provide that as an insight into the ways to talk to people. One of the biggest problems we have as researchers is trying to ensure that we’re communicating to voters on their language and not our language, as sometimes political people speak in terms and phrases that are not shared by the average world. In this online community, we can use their language and not our language.

As politics continue to be more divisive and polarizing, it was discussed frequently this last election whether there really are any truly independent voters anymore. Did you see that you could really find independent voters for your research?

I understand the idea that they might be extinct animals, but no, they actually do exist. A lot of people share the view that, like the middle class, they are a shrinking breed. They really exist, and there are a couple of different elements to them. Some of them are apolitical people that don’t like politics at all, and they are independent just because they haven’t put the energy into determining their political views. It’s kind of like, “I’m too busy with my own life to pay attention to politics, and therefore I’m independent.”

Then there’s a second breed of independents that does understand politics, and really just doesn’t like the party system. These may be people ripe for a third party, but may not be happy with politics as usual. Then there’s the “pox on both their houses” crowd, which is kind of related to that second group, that really thinks the Democrats are too liberal and the Republicans are too conservative. Then there’s an offshoot of that even, of people who think both parties are the same; they’re both bought by corporations, and they’re both indebted to big-moneyed interest. They’re independent because they just hate the party system. So all those people exist, but most of them do have a history of voting both for Republicans and Democrats, and they are in play in an election like this.

There are a lot of people in non-swing states who believe their vote doesn’t count. If constituents aren’t already politically motivated, is there incentive to get out and vote for a president?

There is a topic I was just exploring last week with regard to the efforts in some states to change the electoral allocation. There is a bill in Pennsylvania to move from winner-take-all to congressional district awarding. So instead of Obama winning all of Pennsylvania’s electoral votes, it would have been split. Actually the way they’re calculating it, Obama would have lost the state, even though he won the popular vote. The way that the Republicans gerrymandered the state, they control more of the congressional delegation. So awarding upon congressional district would have given more electoral votes to Romney than Obama. They’re actually pushing this legislation, which led me to go out and do a bit of research on how average voters think about electoral reform. Most of them are confused by how the Electoral College works. What makes sense to them is to add up the votes and whoever gets the most wins. That just makes the most common sense. So we no longer have everybody looking at Ohio and nobody looking at Illinois. 

The 2012 election marked a major realignment in the demographics of voters, with a surge of Latinos and youth entering the electorate. How well were you able to see these new realities in the research you conducted leading up to the election?

We did work with that new population a couple of times, primarily about motivation to register to vote, and to figure out how to talk to both youth and Latinos. That was a bit of a struggle. Obviously the campaign did a good job of getting them to come out to the polls, but when we did our research with them, finding an avenue to connect politics to their lives was not an easy task. When we were doing our initial research with them, there were a lot of people removed from politics. I have to say there was a lot of this group that was disappointed with the President’s first four years because they got excited in ’08 about having a new guy that they could attach their hopes to. A lot of them were saying he didn’t deliver. So I give the rest of the campaign a lot of credit for being able to mobilize the new electorate to get them to vote.

Pundits argued during the general election about whether this presidential campaign was a matter of wooing swing voters or energizing the base. From an inside perspective, where did you sense the balance of the Obama campaign?

I think that the balance is relatively equal moving into Election Day, as far as emphasis on persuading the likely voters versus motivating the new voters. My work with the campaign was focused more on persuasion than motivation, but I think as the campaign moved on and the quantitative data came in, it indicated that the number of undecideds was low – lower than usual. People were more dug in to their expectations for which side they were going to be on. The undecideds shrank throughout the course of the campaign, to a very low number. I believe the campaign did become more focused on motivation because they understood the turnout was going to be such a huge factor given that there weren’t that many people to persuade. I think the campaign moved more toward a balance of motivation. My work was still focused slightly more on the persuasion aspect because we were looking at overall messaging. Certainly there were a couple of events that occurred that had the potential of impacting a wide number of voters, like the first debate and the 47 percent remark. So there were times where we had to pay attention to the likely voters to make sure that they weren’t shifting because of those events in the campaign. I do think that overall the motivation of the new electorate was essential.

With your focus on qualitative research, how do you effectively collect and organize qualitative data to analyze and report?

There are a couple of steps we take to do the analysis. A lot of it is just listening to voters to determine common reactions to events. The literal way we do this is, we have a note taker in the back room who is on a laptop, typing up what voters are saying in real time. I, as a moderator, am sitting in the room taking notes for any time that voters say something that’s of interest to me or insightful in anyway. After the group, the first thing we’ll do is write a very quick email out to the campaign that summarizes the important things that we heard. If there were a couple specific questions that week about an issue, whether it’s immigration or Obamacare, we write an email out that night about what we heard. Then there is a longer process that takes two or three days of going back. We sometimes get transcripts, go back and review what people said, take verbatim quotations and straight points that they were making. We’ll put that out in a longer report that tends to both summarize what we heard, with quotations, and also put it in larger perspective, comparing it to what we’ve been hearing before or talking about if it was new stuff. A lot of that is judgment and analytical conservation from us as researchers.

It’s a lot more subjective than polling where you get a report, you run statistics on it, you do testing, and you have answers. You know, you get a certain percent that you put out in a report, with subgroups, and a nice table showing men are different than women, Latinos are different than whites. None of that is applicable to the work on the qualitative side because you can’t have the statistical precision. A lot of it is judgmental and in the mind of the team. Anyone can look at a data set and report back what the numbers are, but the stuff that we analyze is a bit more subjective.

Were your reports used more to make real-time decisions or to direct future advertisements and talking points? 

Both. The information that we provided to the campaign went into the talking points that we delivered to the spokespeople. Some of it would be internal campaign staff, and I believe they even disseminated it out to the non-campaign surrogates who were going out into the press. When we talk about a finding, we send it out to the communications people, and they send it out both internally and externally so people have talking points.

There was equal, if not more, done of that with advertising. It was absolutely a focal point of the use of our research. We also did ad testing other than just talking to people about issues. The qualitative reports that we did were used to help design advertising, as far as messaging within the ads, whether it’s a couple points they were making or icons they put on the screen. They sometimes get information on that from our qualitative research, but also just general messaging, like what do the advertisements have to say. There was another component of us reporting back as to how ads were performing. We did that as well for ads against us, where we reported back to the campaign, “this ad by the Super PACs is very damaging,” or, “this ad from the Super PACs is ineffectual and voters are shrugging it off.”

What were the main policy issues that you found resonated strongest with focus groups?

The most resonate message was how to grow the economy in the future. That has been first and foremost for years now. Looking into the 2012 election, the challenge of the Obama campaign was to move voters from dwelling on the disappointments in the lack of economic improvement during the first term, which is where they came into the room, to examining a choice for the future between two visions of economic future. The President’s vision was more of how do we help the middle class grow, and the contrast that we were trying to provide was that the opponent’s model for growing the economy was one that had been previously tried, failed, and was not the way to go. What we were listening for was the degree to which voters were saying, “I don’t know how the Republican model of less regulation and less spending is going to help me and my family.” That was the focal point of what we were looking at. There were other issues of course: health care and war. Women’s health was extremely important, but the economy was foremost.

Feature photo: cc/(goir, photo ID: 77475635, from iStock by Getty Images)