Interview with Professor Michael G. Miller

CPR’s Dimitrius Keeler asks Professor Michael G. Miller about election forecasting, battleground states and the future of American politics.

Dimitrius Keeler, CPR: Thank you for agreeing to this interview. Let’s jump right in. Do you think that the recent report about Trump’s tax returns will be a problem for his campaign, or is it more likely to blow over like the other scandals we’ve seen this year?

Professor Michael G. Miller, Political Science, Barnard College: As I see it, this will not cause much of a shift. If you are in the Trump base, this revelation is pretty easily spun. You can say that he has exploited the tax loopholes available to him, which is the sign of a good businessman. That’s certainly the line that the Trump campaign has taken. On the other side, if you’re a Clinton supporter you can say, “Ah ha! I knew it, he doesn’t pay any taxes, and it’s utterly consistent with what I thought.” So regarding the shift of public opinion, I don’t think we’re going to see a lot of that.

I approach this exercise as a campaign would: I would look at the baseline registration numbers because those numbers preview how the county is trending.

But the scandal is going to consume the most crucial resource that Trump has left, which is time. He will get at least a week’s news cycle on this. He’s going to have to do a lot of explaining because taxes are a very complicated issue. I think we need to stress that he’s done nothing illegal here, but most people don’t understand the nuances of the real estate tax code. There’s an adage in politics that if you’re explaining, you’re losing. This scandal is not going to help him during an already bad month for his campaign.

DK: As someone who has worked on campaigns in the past, what are your thoughts on Trump’s campaign structure, putting so little emphasis on the field organizing on the ground and fundraising? Does this change the general thinking about campaign strategy, or is it an anomaly that we’re unlikely to see again?

MM: Most political scientists would agree that it shouldn’t be replicated again. When we forecast elections like this, we always start with what we call the fundamentals, which are the pieces on the board that the campaigns really can’t change. It’s just the way the game is set up. Those are things like the economy and the popularity of the current president. If anything, those fundamentals really favor Secretary Clinton. Everything we have seen throughout this election demonstrates to me that Trump’s unwillingness to engage in traditional campaigning has really hurt him. If we look at the number of offices that the Clinton campaign has in battleground states, she has a massive advantage over Trump. We know empirically that the best way to get people out to vote is to interact with them in a high-quality, close way. Trump’s ability to do that is really compromised by his strategy.

Secondly, his media strategy has been unlike anything we’ve seen in a modern campaign, in that he goes in fits and starts. He’ll be up with ads for two weeks and then he’ll go completely dark. That’s just not the way this should be done.

DK: Have you seen any improvement in Trump’s campaign as he’s cycled through campaign managers, going from Corey Lewandowski, to Paul Manafort, and now Kellyanne Conway?

MM: A major problem is that there are pretty clear factions within his campaign. One of the challenges he’s had as a candidate is that he seems to have these warring tribes in his own organization, and he’s been unable to unify his own people. That’s going to result in different advice. For example, as a student, if you go to the professor and the TA and get different pieces of advice, it’s unclear who you should follow. Trump, a fairly inexperienced candidate, has proven himself to be susceptible to that.

There’s something to be said about experience. The President of the United States seems to be the only job in America where people are saying that the least experienced person is the most qualified for the job.

DK: Can you speak to what you’re seeing in the electoral map and the polls coming out of various different states? You have been analyzing specific voter registration figures coming out of several Pennsylvania counties. Are you seeing anything in the map that you find particularly noteworthy?

MM: I like to explain the map from Trump’s perspective: to win this race Trump needs to win Florida, Ohio, and Pennsylvania. It will be almost impossible for him to get to to the White House without those three states. This puts Clinton in an advantageous position because all she really has to do is hold one of these and she will likely win the election. Two weeks ago, things were trending very well for Trump in Florida and Ohio, but now Clinton seems to have a lead in Florida. If the race ended today I think Trump would probably win in Ohio. In this case, if you’ll excuse the pun, Pennsylvania is the keystone state. I think it’s the keystone of this election.

It’s also a very interesting state from the perspective of political democracy because you have these three major urban centers—Philadelphia, Pittsburgh, and to a lesser extent, Erie—where Democrats really have to improve their performance. If you’re watching the election on that Tuesday night, and one candidate is called for Pennsylvania, I think you can go to bed. If we acknowledge that Pennsylvania is that important to the outcome of the election, then it is useful to get to know Pennsylvania on a sub-state level, down to the county level.

There is an adage among political consultants who specialize in Pennsylvania: in the west, it’s Pittsburgh, in the east, it’s Philadelphia, and the middle is Kentucky. This is an oversimplification, but it does accurately describe the voting demographics of the state. You can reasonably expect Donald Trump to win maybe 90 percent of the counties in Pennsylvania. Clinton might take 10. She has definite targets that she has to meet in Philadelphia County, in Pittsburgh, and in Pittsburgh’s suburbs. Allegheny County, where Pittsburgh is, has more Republicans than any other county in the state so that will be be a real fight. If Clinton can win Allegheny County by roughly 53 percent, that sets her up well, but she also needs to get people to vote in the city of Philadelphia. It can be useful to look at Bucks County, a suburban Philadelphia county that has been very close in the past and will likely also be close in this race, as a barometer for how the race might go in Pennsylvania at large. 

I am also interested in registration totals, because they tell me as a forecaster what I can expect when I’m setting up baselines. Voter registration is also the first thing a campaign will look at. I differ in this respect from other political scientists who focus more on polls. I approach this exercise as a campaign would: I look at the baseline registration numbers because those numbers preview how the county is trending. Journalists have been writing stories that suggest that Republicans have higher voter registration this time. But actually, in places like Luzerne County, Pennsylvania, while Republicans are registering more voters, they are primarily people who used to be Democrats and who have been voting like Republicans in the last several elections and are only now making it official. There is not a lot of evidence that Democratic voters are going away and being replaced by Republican voters. In fact, the evidence in many parts of the country suggests a Democratic advantage in registration, particularly in important places such as Colorado and Philadelphia County. Ultimately, these registration numbers yield no clear conclusion, and the whole election comes down to ten counties.

DK: Ten?

MM: Ten counties in the United States will likely determine the outcome of this race. It’s not an exercise in persuasion, it’s an exercise in mobilization. So which campaign is better equipped to identify what those ten counties are and mobilize people in them? That’s how this works, and that’s why those field offices are so important. In addition to the counties that I have mentioned, Allegheny and Bucks in Pennsylvania, there is also Hamilton County in Ohio. If Lake County goes for Clinton, she wins Ohio, and she wins the election—it’s that simple. Within all of these states we can find these battleground counties and start forecasting how a state is likely to go based on the results in these counties.

DK: How do media attention and debates contribute to efforts to locate, persuade, and mobilize voters, and how do these methods compare to campaign groundwork?

MM: The high-level media can direct voter attention to the broader chess board. The real role and power of campaigns lies in pointing voters to the fundamentals. Voters don’t always have a good sense of how the economy is performing. If the campaign can clarify the broader state of the economy, they can connect the economy to preference. It thus becomes an exercise in prognostication. Campaigns are actually pretty sophisticated at assigning correct forecasts to political candidates, and they use that information to persuade voters.

However, the act of mobilization is different from persuasion. The Clinton campaign is in the final stages of what’s called voter ID, which means that they’re using the voter file and the consumer file to predict how voters will behave, and then they’re trying to reach the voters they think will be supporting them to confirm their predictions. Once every voter gets a grade, the campaign moves to the second phase: get out the vote. This starts happening when early voting starts. Campaigns have legions of volunteers contacting known supporters to make sure they vote. That’s how mobilization is accomplished: people standing on doorsteps, people making the calls.

DK: A few months ago it seemed like a third party candidate could have an interesting impact on this race, but after Johnson’s recent gaffes, it seems like a lot of that excitement has died down. Is there anything still to be said about third parties in this election?

MM: Third parties are interesting, but most of my colleagues and I feel they are overblown in the media. The way that we elect people to the federal government and the state government mathematically guarantees that we will always have two parties.

DK: This is called Duverger’s law, right?

MM: Correct. Voters innately know this, but the Libertarian voter faces a choice. You can either vote your conscience—like John Quincy Adams said—and cast the vote that you know is most reflective of your inner soul, or you can vote strategically, and cast a vote for the party closest to your beliefs who also has the highest chance of winning. Once aggregated, this dynamic effectively assures that only two parties will be viable. To the extent that third parties might affect the race, it depends on how persuasive they have been at convincing voters to make protest votes.

Then it’s a question of which candidate is losing votes to third party protest candidates. For this election, it’s not clear. In 2000, we definitely knew that Al Gore would be hurt by the presence of Ralph Nader. Today I think you can make a case that both candidates will lose votes to Libertarians. I have some theories about this, but for example I don’t see Gary Johnson winning electoral votes. He may get above five percent of the vote, which would be a huge win for a minor party, but that’s about it.

DK: It’s hard to think of a scenario in which Bernie supporters, disillusioned with Secretary Clinton, would support a Libertarian candidate whose beliefs are so incongruous with their own.

MM: I’m not sure that I agree that Johnson is the polar opposite of Bernie Sanders. According to Libertarians, the appeal of the party is that every American is a libertarian. This is half correct. Whether you are liberal or conservative, you can point to elements of the Libertarian platform that you find attractive, so that Bernie voters may have Libertarian views on the so-called “social issues,” particularly on things like marijuana or sexual behavior. If Bernie supporters prioritize those issues over economic issues, they could defensibly move in that direction. However, the larger problem is that we have reached a point where we are making unreasonable demands on our candidates.

I think the some of the resistance to Clinton has emerged from the Obama years, in that the electorate now demands inspiration from candidates. Potential Clinton folks are not finding her inspiring, and they’re making up reasons for being unable to support her, and they’re going elsewhere. I don’t have any empirical evidence for that, but it’s a working theory that I’ve begun to study. I tell students that voting should take five seconds, and it should be devoid of all feelings. Voters should think that this is the person who will advocate the policies that they think are best. We don’t need personal connection to the candidate beyond that, but we have recently begun to demand one. [Laughs] In this way, I guess Barack Obama ruined voting in the United States.

DK: According to the Rational Actor Theory, voters choose candidates who most closely represents their beliefs. How might voters’ motivations change to account for their non-rational feelings and emotions?

MM: Everybody comes to voting decisions differently. For example, I don’t demand to be inspired. I take a cold look as a voter at what the candidates say they’re going to do, and weigh that against how much I trust that they’ll do it. But everyone comes to his conclusion differently. There are people who will vote strategically. There are also people who will prioritize a single issue. There are a lot of ways to reach a voting decision. 

DK: Considering your expertise on campaign finance, can you discuss the potential constitutional amendment to repeal Citizens United? What have the effects of Citizens United been from an empirical standpoint, and does the decision draw more attention and admonishment than it deserves?

MM: I think it does get more admonishment than it deserves. It’s difficult to observe obviously corrupting effects. You can reasonably argue that all money is corrupting, and since now there’s more of it, candidates will feel beholden to the people who funded their super PACs—but maybe that is impossible to prove. So that’s the basis of that claim.

There is much more money in politics today than there was ten years ago, and you can trace that directly back to the Citizens United opinion. The Supreme Court’s position rules that there is no problem with more money in elections because it represents more political speech feeding the marketplace of ideas. There is no evidence—or I should say, there is no evidence yet—that super PACs are buying elections, or that candidates are breaking the law in a systematic fashion, but there are reasons for us to be concerned.

The cornerstone of Justice Kennedy’s opinion in Citizens United rests on the assumption that super PACs could function without coordinating with candidates. We are beginning to see some signs that the Justice’s supposition was probably naïve and that this is a problem that will need to be tackled. In my opinion, Citizens United has actually had more consequences in state elections because in the states there is a much wider range of campaign finance regulations and some states were actively involved in public election funding. Citizens United really frustrates those proactive reforms; it has shaken up elections and campaign finance environments in some states to the point that we’ve seen states scramble to try to to keep up with these opinions.

DK: While we’re on the subject of state elections, I’m interested to hear your take on the potential negative effect of Trump’s candidacy on down-ballot races. In congressional races, are we more likely to see Republicans voters who are crossing over to vote for Clinton split their ticket to vote for Republican legislators? What’s the precedent for this?

MM: Empirically, I don’t know what to expect. I haven’t really delved into polls, and I’m not sure we have good data on who people are going to vote for at the top of the ticket, compared to who they are going to vote for at the bottom until after the election. Paul Ryan or Mitch McConnell are trying to manage elections in their congressional conferences, and I honestly think, in their heart of hearts, they don’t care who wins this election. I think they both recognize that they are going to have a difficult opponent in the White House regardless.

What I think keeps them up at night is this question of whether Republicans are going to vote for Clinton or Johnson and then vote a straight Republican ticket down ballot, or whether there will be people like Jeb Bush or other typically reliable Republicans who say, “I’m just not going to go out this time.”

That’s really what these legislative leaders are concerned about. It makes congressional races much more difficult to forecast, and I think we honestly have no idea what’s actually going to happen in these races. If the Democrats take the Senate back, which seems slightly more likely than a coin toss, I think we could probably credit Trump for that. I was originally interested in whether the damage would be so bad that it could threaten the Republicans’ House majority, but that does not seem to be likely. 

DK: Are there any Senate races in particular that you’re looking at as ones we should be paying attention to?

MM: Indiana and Ohio are both really interesting. I think in Ohio, the Senate race is going to be just as close as the Presidential one. You have two very experienced politicians there. In the Indiana race, the Democrats got a boost with Evan Bayh on the ballot. Actually, I would add that those races are interesting for different reasons. In Ohio, you have two of the most seasoned politicians in the state running for the Senate seat. In Indiana, you have an experienced insurgent Democrat, and in Illinois, a liberal state, you have Republican Senator Mark Kirk running against a sitting Democratic member of Congress. All three of those are seen as really strong pick-up opportunities for the Democrats, but in different ways. Those are my favorite races this election season. 

DK: You alluded to Paul Ryan’s sentiments towards Trump, and to the anger of establishment Republicans who don’t see themselves represented by any of the candidates. How do you see the future of the Republican Party after this election given the deep factions in its ranks?

MM: It all depends on who wins. Honestly, if Hillary Clinton wins, the probability of things reverting to business as usual for the Republicans is high, because this is a party that is quite strong in opposition. Both the Republican Party and its electorate unified in opposition to a Democratic president. I think there is a higher probability that the so-called mainstream establishment will regain control of its machinery in that scenario. If Trump wins, it will provide validation for the populist elements of his base. For instance, you’re probably going to see some new faces in Cabinet positions. At that point, with a Republican president who would disagree with the Republican Speaker of the House on a lot of policies, the Republican Party would have to look in the mirror and decide what it is.

Somewhat paradoxically, if Donald Trump wins this election, I think it will at least spell a serious moment for the Republican Party, if not its demise. There is historical precedent for this. At times in American history, we’ve seen minor parties, such as the Libertarians today, waiting in the wings and looking for a factious split in a majority party so that they can come in to fill the gaps. That scenario would look like the establishment Republicans throwing up their hands and saying, “We can’t be part of this party anymore, it doesn’t represent the party that we joined,” and then creating some kind of alliance with the Libertarians to form a new party. This would arguably pull the Republicans back to the pre-Reagan party that it once was: focused on business, and not on social issues. This would leave the Trump Republican brand out in the cold. Of course, that’s not a prediction, but just a scenario. In the 1840s, the Whig Party’s election of Zachary Taylor offers a similar situation. The Whigs were completely divided by factions over the question of slavery, and were unable to move forward. After that, Republicans filled the gap. This birth of the Republican Party could foreshadow its death.

DK: You teach a class on women in American politics. Given that a woman has earned the nomination of a major party for the first time , how do you feel about the media’s coverage of Secretary Clinton compared to that of her male predecessors?

MM: I think the media is doing better than they did last time. We have a lot of evidence that the 2008 coverage of Clinton was much more negative than that of Obama. We are well aware of the double standards that women face as candidates. They are much more likely to face commentary on their appearance, demeanor, and emotions than men, and that’s a hard headwind to run into. I think she has been bolstered significantly by the experience of campaigning before. I’m happy to see fewer gender comments this time, but to say that she doesn’t face unique challenges would be factually incorrect.

DK: I’m curious to hear what your personal experience has been covering this election as both an academic and a political forecaster, compared to your work on previous elections.

The real question is where my resbonsiblity as a teacher lies. If
I think Trump is a dangerous person, is it my responsibility to communicate that?

MM: This election is much harder to look at because there are so many shifts going on. One broad problem is that we don’t have a good understanding of Trump’s base, and if it will vote—that is really our biggest problem right now. We won’t understand the “Trump Voter” until after the election when we have better data, which really affects the polls. Every poll that we look at before the election is using a likely voter model to try to determine if the people we are talking to will actually vote. This is very hard to do for new or non-habitual voters, and Trump is definitely drawing his base from some of those people. Combine that with his apparent immunity to all of those moments that would be game-enders for just about any other candidate. Frankly, many of us have been wrong about Trump at every juncture of this race, including me, which makes me reluctant to make a prediction. In a typical race, I could confidently predict the outcome by now, but I’m delaying until at least two weeks out from the election. I’m not sure that I will trust my own prediction until then, since there’s so much in flux.

DK: There’s an expectation for professors to maintain some semblance of neutrality in politics, but I imagine that has been difficult this election. Has this election impacted your teaching?

MM: This has been a difficult pedagogical election, and this has been a conversation that many of us have had online, on forums and on social media. Many of us prefer a government with two strong, viable parties. I think we get better policy when Republicans and Democrats are strong, focused, and in a position to bargain. It’s not clear to many of us what Donald Trump is. Is he a conservative? Is he a populist? Is he a neo-fascist? I think you can make arguments for all of those labels. The real question is where my responsibility as a teacher lies. If I think Trump is a dangerous person, is it my responsibility to communicate that? That’s a real question that many of us have been wrestling with. Furthermore, can you divorce Trump from the statements made about the party itself? What I have been trying to do is separate Donald Trump and any statements I might make about him as an individual from statements about the merits of the Republican Party of politicians like Paul Ryan.

DK: That seems like a really difficult split to make.

MM: It’s tricky to get students who may not otherwise care about politics to understand that the locus of power in the Republican Party still rests with the Speaker of the House. When I talk about Republican policy, I’m not picturing Donald Trump, I’m picturing Paul Ryan. In no other election would this be the case, and I think that says a lot about the state of the Republican Party.

DK: I know you’re running a “Most Influential Non-Politician” tournament in your class. Do you have any early forecasts, or are the winners unclear?

MM: Everyone should vote! The winner will surely be Martin Luther King, Jr., right? It must be him. But there are some interesting match-ups coming. We have Malcolm X versus Robert E. Lee in the same bracket, for instance, so there will be lots of interesting conversations. We have 96 people in the bracket, and I put in a lot of people whose names are not familiar to students, so I can force them to think about such lesser-known figures. I’m really excited about it.