Editor’s note: This interview was conducted prior to the announcement that former U.S. Senator Rick Santorum would suspend his campaign.
If you want to know what separated the supporters of the two frontrunner Republican presidential candidates during the primary campaign, political scientist John Lapinski says to look no further than religion.
In the volatile and prolonged Republican primaries, religion emerged as a noteworthy voter trend that distinguished supporters of former U.S. Senator Rick Santorum (who suspended his campaign on April 10) from those who backed former Massachusetts Governor Mitt Romney.
“It is really the evangelical vote that’s carrying the vote, not so much the Catholic vote,” says Lapinski, associate professor of political science in the School of Arts and Sciences. “A lot of people thought it was going to be he Catholic vote that was going to help [Santorum], but he [didn't] do as well as one would expect with Catholics.”
As a political scientist who has taught and written about Congressional lawmaking, American political development, and elections, it’s only natural that Lapinski watched the 2012 Republican primary season with great interest.
There’s another reason, too. Lapinski works for NBC News’ Elections Unit, as part of the Decisions Desk. It’s an important gig, considering Lapinski leads a team that helps make the call about projected winners in primaries, caucuses, and general elections. Every time you heard on NBC, “Mitt Romney has won the state of Illinois,” or “Senator Santorum has won the states of Alabama and Mississippi,” that’s, in part, because Lapinski and his team have analyzed statistical models, complicated logarithms, and data from polls in a quarantined room.
It’s a job that Lapinski has done since 2000, when he got involved with the network, he says, entirely “by accident.” He was interested in survey research, and contacted then-MSNBC editor-in-chief Merrill Brown to explore the idea of doing an online survey experiment with the news website. Afterward, Lapinski was offered the opportunity to work in NBC’s Elections Unit as a consultant.
The Current sat down with Lapinski in his office to talk about the busy Republican primary season, how exit polls help his television team determine projected winners of races, and if voters’ opinions about politicians really change over time.
Q. Was this election cycle actually different than other ones in recent memory?
A. A lot of people think that every election is unique. There are things that make this particular election different than others, but there are also some things that make it similar. In 2008, it was a different primary on the Democratic side. For one thing, caucuses became extremely important. President [Barack] Obama exceeded Senator [Hillary] Clinton in the caucuses and that’s how he built what eventually became an insurmountable lead. Caucuses differ from primaries in lots of ways, but one big difference is that they’re harder to participate in and they’re less representative. Now this time, there’s almost been an overemphasis in the sense that we’ve spent a lot of time looking at caucuses.
One reason is, like 2008, it [the primary season went] on for a very long time and one of the reasons is the rules have changed. On the Republican side, the rules had always been structured in such a way that they [would] pick nominees quickly. In the primaries and caucuses, it was winner-take-all. … The Republicans changed that process because they thought that the nominee was picked too quickly. [There also have been] changes in campaign finance, namely the Citizens United decision [that] has allowed candidates to receive the assistance of super PACs, and campaigns can go on almost indefinitely, even if they are not raising a lot of money.
Many things have lined up and that’s what’s made this race a little bit different. It’s also because there have been so many players in it. And also, the other thing that might make it seem like we’ve been following it for a long time, is the debates were out of control. I watched every one, and there are multiple reasons why there were so many of them. … There were millions of people tuning into these things.
Q. Has there been any kind of enthusiasm gap among voters who identify themselves as Republicans?
A. Everybody’s talking about that. Especially early on, some people were thinking that maybe there was an enthusiasm gap because people thought that Republican turnout was a little low. While turnout has been low in some places, it’s been high in others. It’s not a universally even story where Republicans turnout is low race after race. Just recently, turnout in the Illinois primary was very solid. Illinois is a state that leans Democrat, but the Republicans turned out. In some of these other states, turnout has been low. It’s hard to tell that story in a coherent way.
One thing that we’re certainly seeing, and we’ve seen in the past but we’ve put more emphasis on this time, is the caucuses have had some trouble this year. In Iowa, the party projected that Mitt Romney won and then all of a sudden, ‘We’ve made a mistake, it’s Rick Santorum.’ There have been other instances in Maine. Some of the votes got thrown into a trash folder, so not all of the votes were initially counted. We’ve seen more issues in the caucuses than in the primaries. With the caucuses, you get these states where there’s very low participation. For example, in Wyoming, you’re talking [about] hundreds of people who actually participated. In Maine, I think it was maybe 5,500 people. … When I was covering the caucuses in Maine, the party chairman came out and said it’s been record turnout—and it was record turnout, but turnout went up by 1 or 200 people. When you’re thinking about that, there are some issues there, when such a small number of people are the ones who are participating in choosing which candidate will win. And also, in states where delegates are allocated or tied to the [caucus] vote, they’re picking their delegates for the nominating convention.
Q. Among Democrats—looking at the hard numbers, is there an enthusiasm gap?
A. The question right now, and the question in the general election, is going to be one of turnout. One of the great stories of the 2008 election was that Obama built a turnout machine. He particularly did well in caucus states where there was a lot of enthusiasm for him, and a number of empirical analyses that show he brought in a lot of new voters into the fold. There were a lot of first-time participators in the 2008 election. The question will be—and this is an open question—what will happen in 2012? Will those people who turned out and voted for him in 2008 turn out and also vote for him in 2012?
We have no idea yet whether that will happen. There are a lot of people talking about that and that’s why they want to talk about the enthusiasm gap, but they have no evidence.
Q. Do you think new voter I.D. laws could hamper turnout in any way?
A. That’s to be known later because I don’t think they matter as much in Republican primary caucuses. I know some of the literature, but it’s not my area of expertise—but it’s definitely something that is more of a partisan issue. Democrats think it’s hurting them, in the sense that people who are more likely to not be able to participate are people who would be more likely to vote Democrat.
Q. In part, you look at bigger trends among the electorate. So, what do you think the electorate wants this year? Is it, in fact, all about the economy?
A. What I’m mostly looking at right now is the Republican electorate. Our group does a number of different things, but we use exit poll data and real vote data to make projections. Exit polls are essentially surveys where people who are participating in Republican contests take an exit poll, or an entrance poll in the case of a caucus. One of the questions that we ask them is, ‘What’s the most important issue facing the nation?’ and one that’s identified as being the most important is the economy. There’s no question about that. But, in this particular contest, there have been a lot of interesting voting trends. One thing that [divided] the race between people who supported Governor Romney and people who supported Senator Santorum is religion. Everybody agrees that the economy is important, but people who consider themselves to be evangelical Christians, or born-again Christians, strongly support Senator Santorum and they [did] not strongly support Governor Romney. There’s been a lot of political geography that’s been brought into this analysis. People say Senator Santorum [did] really well in the South, Governor Romney [did] well on the coasts—and he’s won all the big midwestern states. Iowa was essentially a tie. … Senator Santorum won Alabama and Mississippi on March 13, and a lot of people thought maybe he wasn’t going to win. I thought he was going to win. We had some polling that suggested he wouldn’t win, but my priors said he was going to do well in the South.
Now, of course, there are other demographic factors that have been important, too. Governor Romney is clearly better with the better educated, he does really well with people who make over $100,000 a year. Education and income, there’s been a split there. But really, the religion issue has been pretty important and Senator Santorum brought that up quite a bit in his campaign. That’s interesting.
There are some states that are not particularly competitive, and I can’t see those states not going for whomever the Republican nominee is."
Q. Now that Mitt Romney is the presumptive nominee, do you think the other candidates' voter base will come around to back him? Will evangelical Christians reluctantly embrace Romney?
A. I don’t even know if it would be ‘reluctantly.’ If you were to ask me in a general election if Governor Romney becomes the nominee, which is likely given how many delegates he has. ... I can’t imagine Alabama, Mississippi not supporting Mitt Romney. Now if he’s the nominee will they hold their nose and do it, or will they do it more willingly? That’s hard to know right now. That gets into this whole turnout question, again, in a sense of who’s going to participate, and we certainly don’t have any hard evidence of this but a lot of evangelical voters are not particularly fond of President Obama. There’s been a lot of polling—some of it not very good polling—but I saw some robo-type polls that came out of Mississippi that showed a majority of Republicans who are participating in these polls thought that President Obama was a Muslim. And President Obama has said, time and time again, that he’s a Christian. My guess is that [those voters] would support the Republican nominee, whomever that is.
There are some states that are not particularly competitive, and I can’t see those states not going for whomever the Republican nominee is. It doesn’t really matter in some of these states whether the GOP candidate wins by 10, 20, or 30 points. You don’t need to win by 30 points; you’re [still] going to get all of those Electoral College votes.
Q. What about in the states that Obama carried in 2008 that were previously red—I’m thinking of Virginia and North Carolina. Will they be tough again this time around?
A. They’ll both be competitive battleground states. The demographics for Virginia have been getting better and better for Democrats. North Carolina is going to be harder for him. North Carolina was, in many ways, a surprise win. People thought it was competitive [in 2008], but the fact that the state was carried by the Democrats was a game-changer in some ways. The only way that President Obama can win a state like North Carolina is if he gets great differential turnout. He’ll have to really mobilize young voters.
If President Obama wins North Carolina, it’s going to be a good year for him. … It’s really hard to know this far out exactly which states will be the real toss-ups. North Carolina is one that could be in that category, but it might not be. We won’t know that really until quite a bit later, until after we’ve done more polling to see where things stand.
Q. Will most polling take off this fall, once the general election kicks into high gear?
A. In general, there are a lot of bad polls out there. If you go to one of these websites that aggregates the polls for you, like RealClearPolitics and Pollster.com, what you see are a number of different polls. The problem is a lot of these are really terrible polls. They’re robo-polls where you do not have a live interviewer, they do not call cell phones, sometimes they have very strange sampling procedures, and oftentimes, they don’t [give you any information] about the poll. The reason why there are so many of them is they cost no money to do. I could set one up in my office here. I could get a robo-caller and do it myself for almost no cash, especially if you’re using a voice-over-internet phone system.
That’s in comparison to a high-quality telephone survey; maybe you add cell phones and you make sure that you do things like call people back … you might pay $20,000 to $30,000 for a telephone poll [like that]. It used to be when there were a lot of telephone polls you would see things take off, for example, in September and October in an election year, either presidential or Congressional. Now, what you see is people are doing lots of polling, but again, it’s really hard to know whether to trust the polls. You don’t know if you have a good poll or a really terrible poll, and it’s really hard to know what to make of it before an election. How valuable is it? It’s more valuable than no information, but again, a lot of these polls have been way off.
For example, on February 7, all of those polls were outside of the 95 [percent] confidence interval, which basically means they were way off. Romney, in some of those polls, had a double-digit lead in Colorado, and he ended up losing by a few points. Some of them showed an extremely tight race in Minnesota. Some of them actually showed Romney up in Minnesota. Romney did not do well in Minnesota. And in Missouri, they showed Santorum leading, but not by as much as he actually won. In contests where it’s hard to know who’s going to participate, it’s really difficult for a robo-poll to work, because a robo-poll can’t ask a lot of questions.
If I were polling in a caucus state, the first question I would ask somebody is, ‘Do you know where your caucus location is?’ If somebody couldn’t answer where their caucus location is, I would basically know they weren’t going to participate. Robo-polls don’t do things like that. It’s tricky to ask those types of questions without a live
Now, what you see is people are doing lots of polling, but again, it’s really hard to know whether to trust the polls. You don’t know if you have a good poll or a really terrible poll, and it’s really hard to know what to make of it before an election."
Q. How do you exercise a certain amount of caution when it comes to exit polls? They were famously unreliable in the 2004 election.
A. First of all, exit polls come in in different ways. We want to get a representative picture of what’s going on in a state throughout the entire day. Basically, exit poll interviewers are interviewing people throughout the day and then calling in the results at different periods. You get some results in the morning, some in the afternoon, some before the polls close.
It used to be the case, a long time ago, that [the data] flowed in and everybody at the networks could see this data. The problem with that is, sometimes the results look very skewed at a particular moment. For example, Mitt Romney has done very well with the elderly this time. The elderly tend to vote early, so if you only look at the early exit poll data, you might get a picture that Mitt Romney is doing better.
What the networks have done is essentially quarantine the data where only a few people from each organization that participate in the consortium that collects the exit polls can see that data. It’s usually quarantined until the early evening, and you can’t make these mistakes where people are looking at an incomplete, partial picture of the data and then making a bad inference. And you could easily see how you could make mistakes if you don’t look at the complete picture.
I’m involved in that process. There’s a small group of us, but we have no access to anybody at the network. You’re literally in a quarantine room where you can see that data but you have no access to the outside world at all.
Q. And this is so no one is influenced by reports coming in?
A. The other thing that networks have pledged is not to project races before polls close so you can’t make a projection that somebody’s won. You’re not supposed to project, or even characterize, a race to give away a winner until all the polls have closed in a particular state. That’s critical because you don’t want to dissuade people from participating. If all of a sudden, you project a vote before the polls close, people might say, ‘Well, what is the point in voting if we already know the winner?’ The famous case of that, of course, was California in 1980, where the networks had projected that Ronald Reagan had won before polls closed on the West Coast.
What’s the consequence of something like that? First of all, some people may not participate. It doesn’t matter just for the presidential contest—a lot of down-ballot people like members of Congress won’t get votes. People in lower offices might lose because those people aren’t turning out.
Q. How did you get interested in all of this? Have you always been a political junkie?
A. Really, I was not a political junkie early on in life. My family is apolitical. It’s just something that I really became passionate about and interested in in college. It wasn’t just political science, but I liked the social sciences.
My initial goal was to be a social science teacher at the high school level. A couple of people I worked with at the undergraduate level encouraged me to go on. I did a master’s degree at [the University of] Chicago and I got a lot of encouragement from professors there to continue, so I ended up at Columbia, did my Ph.D., and just progressed. I became more and more interested in it
Q. How fluid are people’s opinions about politicians? Do people’s opinions really change over time?
A. It’s hard to know. You have to put it in the context of a situation. A lot of people have looked at the 2012 Republican contest and think that people are very fickle, but that’s not necessarily true. The reasons why people sometimes come to the conclusion that people are fickle is the polls—there’s a lot of bouncing around in the polls.
Where the inference is incorrect is that a lot of people aren’t following the race, and oftentimes people like myself lose track of that: Even though I’m spending a tremendous amount of time thinking about this and following this, a lot of people are not. It might be that we poll them, but they’re giving you an opinion based on very little information because they really haven’t tuned in yet. That’s one of the reasons why you see a lot of volatility in presidential polls until after the first or second week of September. Once summer is over, and people start going back to school, you really see more stability. In national polls there are only contests going on in a handful of states, so candidates aren’t spending a lot of time there. People are making decisions based on limited information.
Q. Is this an especially exciting time for someone who studies this?
A. It is exciting, but it’s tiring. It’s interesting. It’s different. There have been so many ups and downs in this race that it’s sort of unique, in my experience, in the elections that I’ve covered.
Originally published on April 12, 2012