The 2012 US presidential election: a Moneyball approach

John Sides interviewed by Romesh Vaitilingam, 7 Nov 2012

In late September 2012, John Sides of George Washington University spoke to Romesh Vaitilingam about the 'Moneyball' approach to politics – using systematic economic and polling data (rather than 'narrative' and 'gut feeling') to forecast election winners. The US presidential election result confirms the value of this scientific method. The interview includes discussion of the influence of the state of the economy on voters’ intentions, the role of campaign finance and what might happen to US political gridlock after the election. The interview was recorded in Washington DC.

Listen

Unfortunately the file could not be found.

Open in a pop-up window Open in a pop-up window

Download

Download MP3 File (19.5MB)

a

A

See Also

Sides, John and Lynn Vavreck (2012), The Gamble: Choice and Chance in the 2012 Presidential Election, Princeton University Press.

Transcript

View Transcript

Romesh Vaitilingam: Welcome to Vox Talks, a series of audio interviews with leading researchers from around the world. My name is Romesh Vaitilingam, and today’s interview is with the political scientist Professor John Sides of George Washington University. He and his co-author, Professor Lynn Vavreck of UCLA, are writing a book called The Gamble: Choice and Chance in the 2012 Presidential Election, to be published by Princeton University Press next year. In the meantime they have posted two preview chapters on the press’s website. John and I met at his office in Washington, DC, in late September 2012. I began by asking him to explain his approach to analysing US electoral politics.

John Sides: One analogy for the way I do my research on politics comes from the American sport of baseball. There is a famous book and then a movie called Moneyball; it was about the way in which baseball had moved away from trusting the gut instincts of grizzled veterans who evaluated baseball players to trusting quantitative data that provided more accurate metrics about the value of these players, their impact on the team. It was about this one team in Oakland, California, that adopted this philosophy’s somewhat controversial way. So my goal with the study of elections is to move things as close as possible towards Moneyball, which is to say I want to measure what’s going on, I want to measure the background conditions and the fundamentals that are important in any election, like the economy. I want to measure the news coverage, I want to measure the campaign advertising, all of the things that convey information to the voters about the candidates, and then finally I want to measure the voters themselves, I want to understand why they made the choices they did and what impact, if any, the campaign had. I want to separate that kind of approach from the approach that is based largely on testimonials from campaign consultants, that’s based largely on the daily narrative that the journalists produce which is valuable in terms of telling a story of what’s happening at that time but it’s not necessarily providing you with enough analysis to really get at what’s driving what. So part of this goal in writing a book about the 2012 election is to show that you can take that approach and learn a lot of interesting things, and you can still write about it in a way that’s engaging to a broader audience.

RV: I guess with this election in particular, that many people are calling ‘the economy election’, it seems quite obvious to look at economic variables and to think about what impact they have on the polls and on the ultimate voting decisions of the electorate.

JS: It’s a really interesting race because, unlike many eras, the economic picture is somewhat mixed, and it depends on what indicators you’re looking at as to whether things look pretty good for Obama or they look a little bit better for Romney. Usually all the different indicators – economic growth, unemployment, inflation – they’re all pointing in the same direction, so in 2008 there was really no doubt as to which party was favourite by the economy, but in 2012 that’s different.

Getting a sense of the economy is particularly important because what I see out there is people who look at this set of economic indicators selectively or not very carefully, and they’ll say things like: “Well, the economy’s terrible, Obama should be losing in a landslide. Why is he not losing in a landslide? It must be because Mitt Romney’s a bad candidate.” I think a better reading of the economic fundamentals says that actually it’s probably Obama’s race to win. They’re slightly in his favour, and so the fact that he’s leading now is not so much of a surprise, and that shouldn’t lead us to imagine that Romney’s a terrible candidate. He may not be a good candidate, but I can’t say that based on where the race is relative to the economy. If you can get these kinds of fundamentals right, your expectations about the election will be better calibrated, and you’ll be less likely to wander down baroque, inferential paths of blaming Romney’s personality or all of these other factors which might not actually be truly central to the outcome.

RV: Are there some rules of thumb about the relationship between the state of the economy and the outcome of an election?

JS: There’s no definitive law that would make physicists happy, but one thing we say in the book is that, typically speaking, an incumbent president who’s running for re-election is favoured to win even with very modest levels of economic growth. So you really have to be presiding over a recession or something close to it, as an incumbent or an incumbent party, to be at risk of losing. That’s what made the Republicans endangered in 2008 and that’s what made an incumbent like Jimmy Carter endangered in 1980. But for Obama, he’s presiding over economic growth. It’s not going very quickly and it’s not enough to make headline statistics like the unemployment rate go down very fast, but that may be enough. Incumbent presidents are hard to beat.

RV: In the draft chapters of the book that you’ve put up on the internet, before the actual thing appears next year, you call Obama “the unexpectedly popular president”. Can you explain what you mean by that? Is it something to do with the fact that he hasn’t been all to blame by the voters for the state of the economy in recent years?

JS: I just noticed anecdotally that his poll numbers were remarkably stable for the last couple of years. Even when the economic news didn’t seem to be that good, it was never terrible but I thought that somehow it would be dragging down his poll numbers. So what we did is we looked at the presidential approval data that’s been gathered by pollsters since the Truman administration, so we’re talking about 1948 until 2008. We looked at the relationship between presidential approval and economic factors, and a few other things that are known to predict presidential approval in a systematic way. Based on that model of presidential approval we said: “Okay, if this is the way the world worked from 1948 until 2008, then what would we expect from the Obama presidency beginning in 2009?”

What we found was that, towards the end of 2011 and at the beginning of 2012, the model’s prediction proved to be systematically lower than Obama’s actual approval was – somewhere in the range of three to six [percentage] points. There’s enough uncertainty associated with our predictions that we can’t say for sure that the difference is a truly significant one in a statistical sense, but if it’s an accurate estimate then it’s the difference between being a president whose approval numbers are in the high forties, which puts him on the bubble at that point in time in terms of getting re-elected, and being a president with approval ratings in the lower forties, which at least in modern elections means that you’re more likely to lose than win. We tried to explain that with several different factors, one of which as you mention is this idea of who’s to blame for the way the economy turned down, and I think there are still enough voters out there willing to give Obama a pass knowing that the economic mess he confronted as president was one he inherited, not one that occurred under his watch. More people still blame George W Bush than blame Obama. We think that’s part of it.

Another thing that may be operating to Obama’s favour is the fact that, in this day and age, political partisans in the United States are extremely loyal, much more so than they were maybe 30 or 40 years ago. So Obama’s approval rating among Democrats is extraordinarily high. If you go back to 1979 or 1980, Democratic loyalty to Jimmy Carter was much, much lower. He was losing not only substantial portions of Republicans and independents but he was also losing a lot of Democrats, and Obama just hasn’t been doing that. It may be that we’ve entered an era in which it takes a heck of a lot to dislodge a partisan and lead them to actually disapprove of their party’s president or of their party’s presidential candidate. Ultimately, we think that’s kind of an irony because Obama campaigned in 2008 as a candidate who wanted to rise above these forces of partisanship; he said that there’s no red America and no blue America, there’s only an America. But his poll numbers have been buoyed precisely by the blue America that he didn’t want to represent exclusively.

RV: Let’s talk a little bit about his opponent and how he came to get that position as the nominee, which is subject to the other chapter that you’ve posted at the moment. Give us a flavour of how you see his campaign going; from the perspective of the UK and Europe people think: “How can this guy possibly have a chance of winning? The guy’s a buffoon!” It’s interesting how he came to get the nomination if that is true.

JS: It was an unusual year for the Republican party in the sense that early on, as the candidates were getting in to the race, so throughout 2011, there was no real clear consensus behind any one of them. Often what happens is that party leaders and others will look around and talk to these candidates and vet them and talk to each other and say: “Who do you support? Who should we support?” and by the time the first primary or caucus is held, a large number of party leaders have already endorsed one candidate. And that’s the frontrunning candidate. In 2011 that candidate was certainly Mitt Romney. He had many more endorsements from party leaders than any other candidate had, but a lot of party leaders hadn’t endorsed anybody. As a consequence they were sitting out, waiting, seeing, keeping their powder dry, so that suggested that the race was more wide open. That, I think, also helped facilitate these surges by other candidates like Rick Perry or Herman Cain or Newt Gingrich or Rick Santorum who were able, when the news focused on them, to get enough attention to bring some voters in to their camp. But ultimately that was temporary, and I think Romney’s appeal in the party was always quite a bit broader than theirs.

Sometimes people think that the contemporary Republican party is comprised exclusively of people who are evangelical Christians or Tea Party members or pro-life supporters. In fact there’s a large part of the party that’s not as dogmatic, that doesn’t identify with those groups. It’s probably still conservative, certainly by European standards and maybe even by American standards, but not ideologically committed to the extent that pragmatism or other kinds of factors might not matter at all. Romney’s appeal was to these factions within the party, that are actually larger in numerical terms, and he had the ability to appear the more electable candidate of any of the other ones that were out there. And I think that’s one of the reasons that during Gingrich’s surges and during Santorum’s surges there were very few in the party that were actually jumping on the bandwagon. In fact there were many in the party that came out to criticise Gingrich and Santorum openly, or at least to express reservations about what life would be like if they were at the top of the ticket.

So ultimately Republicans, even if they weren’t willing to coalesce behind Romney really early on in the race, did so not necessarily begrudgingly at all. They did so largely willingly, and I think the best evidence for that was that by April 2012, so a few weeks after Rick Santorum had dropped out and Romney was the de facto nominee, 90% of Republicans were already reporting that they were going to vote for him for president. So in some sense all the concerns about a divided party were out of the window pretty quick because Republicans were coalescing behind him, if for no other reason than they want to get Obama out of office.

RV: You said yours is the Moneyball approach to political analysis; let’s talk a bit more about money and the role of campaign finance, because this is something that non-Americans find it hard to get their heads around. From what I understand, Obama was very successful because he raised a huge amount of money back in 2008 and a lot of it was of a different style. It wasn’t raising huge amounts of money from big contributors, it was a lot of small donations by a lot more people. I gather that Romney’s had a lot of money behind him, and I suppose that’s now shifting away as the polls go against him.

JS: It’s a good question. Obama has been raising a lot of money personally for his campaign. Romney has been doing that, but doing less, raising less for his personal campaign. Romney has been supported by a large amount of spending that’s been done in coordination with the Republican party. The Democratic party’s been raising money too, but not in quite as large an amount. Romney’s been supported in particular by independent groups that can raise and spend money to support a candidate, but they cannot coordinate their efforts with that candidate in any way. There are a number of these groups, they’ve existed in many forms in American politics for a long time but they’ve come to the fore in the last two election cycles. By and large those groups are much more well funded on the Republican side than on the Democratic side.

I think that initially people feared that those groups would be able to raise so much money that they could raise unlimited amounts, so you’d have millionaires and billionaires giving huge sums of money to these groups that they cannot give to the candidates under current law. They thought that these groups were going to bury the landscape. It hasn’t quite panned out that way, and there are a couple of things that work. One is that Obama’s campaign message might be strengthened by the fact that he’s doing his own fundraising, and in some sense he’s able to control how that money is being spent. He’s able to control the message. If you’re dependent on independent groups to raise money and spend it on your behalf then you’re also making their decisions as opposed to your decisions.

The other thing is that candidates actually get discounts for the advertising that they buy, as a way to facilitate the ability of candidates to reach voters and inform voters even though they have to pay for it themselves, as opposed to elections in many other countries. Independent groups don’t have those same benefits, so they might be spending five or six times what a candidate spends for the exact same advertising time in the exact same place at the exact same time of day. Ultimately for the last few weeks, somewhat surprisingly, the Democrats have been out-advertising the Republicans in the presidential race, combining Obama’s spending, Democratic spending and independent groups’ spending against Romney, the Republicans and their independent groups.

We’ve still got six or seven weeks to go and that could change between now and then, and it’s not a bad idea to husband your resources and spend them at the end when their impact might be greater in the sense that the impact of the ads will be felt and it will wear off by the time the election comes about. But at this stage of the game I don’t think that the worst scenarios of people who fear the role of money in this election have come to pass.

RV: We hear a lot about the importance of swing states but I think from outside America it’s hard to get a sense of what the micro nature of the race is – what’s going on in Ohio, what’s going on in Florida, what’s going on in Colorado – we just get the top line of “he’s leading by this much nationally”. Can you give us a flavour of how important those races at state level, perhaps even sub-state level, are in terms of influencing the ultimate result?

JS: The candidates are spending almost all of their time and money in a handful of states that we call battleground states. They’re the states where the outcome is enough in doubt that this kind of spending might actually be consequential. Other states – California, Texas, New York – these are big states, they’re important in the election but they’re not really in doubt. We know that Democrats are going to win in California and New York; Republicans are going to win in Texas. So there’s no use wasting your money to try to change that outcome. So candidates spend all of their time and money buying advertising time in these [battleground] states, and they’re doing it in ways that play to these sub-state dynamics that you said. They’re trying to target media markets within these states, and then within these media markets they’re trying to locate their advertising during programming that they think will reach particular groups of voters.

So the targeting extends way beyond just states writ large; they’re really trying to focus the microscope even smaller. On top of that they’re doing much more precisely targeted communications in these states with mail, with contacts with voters from volunteers and from the campaign, and there they can really target individual voters much more cleanly than they can with television advertising. All the dynamics of the races are really taking place within these battleground states. The one thing about the American elections, though, is that the trends in polls nationwide are often pretty well reflected in trends in the polls in these states. If something were to happen in the campaign, a big dramatic event that really changes the national poll numbers – it’s unlikely that something like that’s going to happen now, but if it did – then the same thing would happen in the states eventually.

So, to some extent, as the nation goes then so go the states. But there’s enough wiggle room in these individual states that candidates believe that their campaigning might make a difference on the margins.

RV: What are the kinds of things that could influence the outcomes of the race right now? You’ve said the fundamentals basically favour Obama but you also talk about specifics, so are there some specific things that could happen? We’re just about to go in to debate season, might that make a difference? Might there be some kind of October or November surprise, or whatever you call these big foreign policy incidents that throw things off at the last minute?

JS: The presidential race at this point in time, historically speaking, doesn’t change a whole lot. The biggest events in a presidential race are actually the party conventions, and those tend to produce the widest swings in the polls and they did produce some swings this year, in particular for Obama. After the convention season is over the main event that’s going to take place is the debates, as you said. Historically speaking, you can identify movement in the polls before and after the debate on the order of two, maybe three, sometimes four points. But it’s rare that that movement is sufficient to change the dynamic of a race wholesale. If this were a race that was similar to 1960 or 2000, in which the candidates were virtually tied, there two points in a debate is worth something. But I think probably more likely in 2012 is that the debates, and maybe some of the other campaign activity between now and election day, will narrow Obama’s margin over Romney but not ultimately close it. It’s just relatively rare for an underdog to come from behind precisely because of the debates.

The other kinds of events that could take place, the proverbial October surprise, I think that’s more of a mythical creature because it’s hard to find an American presidential race that was really upended by a late breaking event. Those kinds of events might have a marginal impact, for example some people thought in 2000 that the polls tightened towards the end because of revelations about Bush’s conviction for driving under the influence when he was a much younger man. That might matter a little bit at the margins, but the unfortunate thing for the Romney campaign is that history suggests that the state of the race now is likely to be pretty similar to the state of the race come election day.

RV: You’ve alluded to the influence of the growing polarisation of American society and the polarisation of American politics and I wonder whether, looking beyond the election, whether there are going to be any changes. Are we going to see any breakthrough in the gridlock that seems to be holding the nation to ransom?

JS: I think the basic dynamics of polarisation and gridlock are not going to change. For one I think we’re going to have a divided government. If you assume that Obama’s going to win and the Republicans will retain the House of Representatives –the Senate is in some doubt – but no matter whether the Republicans take control of the Senate or they don’t I think ultimately there’s going to be a divided government. It’s going to force compromise, and of course it’s difficult to force compromise in this day and age when the political parties don’t see eye to eye on very many issues, they’re increasingly divergent. This is not just a story about them not wanting to be nice to one another or being mean-spirited, it’s actually just honest disagreements about the major issues in American politics like the role of government. It’s tough to bridge those differences when you’re that far apart. Obama’s been saying things like he thinks the election’s going to send a signal that makes the Republicans compromise with them. I don’t think that’s true, I think Republicans are going to interpret this election in a variety of ways but none of them mean that they should necessarily compromise with Obama; they could easily just say to themselves: “Well, if we hadn’t nominated Romney we would’ve won,” or a thousand other things. If anything is going to make some degree of compromise, albeit temporary compromise, it’s the fact that there’s a lot of major decisions that have to get made relatively soon where I think neither party really wants the outcome that would result in no decisions made.

One example is that there were a set of tax cuts passed under the Bush administration in 2001 that are due to expire. Once those expire then taxes go back up, and I don’t think any Democrat or Republican necessarily wants that to happen in exactly that way. For example the Republicans don’t want them to expire at all, and Obama doesn’t want them to expire unless you’re making $250,000 or more a year is what he’s said. So if the parties can’t get anything done then I think the result is something worse. That may force them to have to negotiate, and they’ll have a few months in what’s called the lame duck session between the election of the president and when Congress goes on recess for the Christmas holidays, and then the next Congress comes in after that point. So the lame duck session’s going to be a busy few weeks, it seems.

RV: Final question. Just looking even further out, do you think that the Democratic victory in this election will in some way cement a very positive future for the party? There are some people that argue that perhaps there’s some truth in the 47% being a Democratic majority. – the growing Latino proportion of the population which seem to be very much in the Democratic camp now – do you see trends that way or can you not look that far out?

JS: I’m really reluctant to make big predictions, and I’ll say why. After 2004 people were saying that Bush had executed a real realignment, that we were headed for Republican party dominance, that his main political adviser Karl Rove was some kind of svengali or genius. Then after 2006 and 2008, when the Democrats took control again, people said: “Oh, now it’s the Democrats’ turn, now we’re going to have a real realignment, this is what we always thought was going to be true because of a larger number of Latinos in the United States growing demographic groups that support the Democratic party.” And then look what happened in 2010.

So I think these predictions of realignment have a relatively short shelf life, and I’m not prepared to suggest that a victory in 2012’s going to solidify anything. Obviously there are some demographic changes ongoing in the country, particularly the decrease in the white population and the rise in the population that identifies as Latino or perhaps as Asian. I think both parties are going to continue to work to engage and to mobilise those groups in their favour, but I think we’re a number of years away from saying that we’re entering into any kind of permanent majority. Every time people suggest that we’re at the end of history, as it were, to use the phrase from Fukuyama’s book, it turns out that history has a few things up its sleeve.

RV: John Sides, thank you very much.

Topics: Politics and economics
Tags:

Associate Professor at the Department of Political Science, George Washington University

Most Read