In the last presidential election, 58.6 percent of eligible voters went to the polls. That’s lower than most developed countries, even after factoring out those, like Belgium and Turkey, where voting is compulsory.
In the 19th century, however, turnout in presidential elections frequently topped 80 percent. Why, more than 100 years ago, at a time when much of the U.S. population was poor, less educated and often lived far from polling places, was turnout so high? Much of it had to do with that century’s get-out-the-vote effort. Illiterate voters simply submitted ballots that ward bosses filled out for them. In many towns, most of the polling places were in saloons, where voters were guaranteed free drinks.
Steering voters to the polls today—what political pros call GOTV, for “get out the vote”—is an entirely different enterprise, one that political scientist Donald Green has made a career of studying, first at Yale University and, since 2011, at Columbia. In dozens of studies and two books, he has looked at the efficacy of virtually every tool at a politician’s disposal, both high tech and low: lawn signs, robo calls, prepaid postage on campaign mailers, endorsements, door-to-door canvassing, to name just some of them.
As Election Day draws near, however, the importance of grand strategy tends to fade. “There comes a point in late September and early October in which it is too late to build a coalition, too late to fundraise to any appreciable degree,” said Green. “At that point, the strategies are set, the economy is what it is, any foreign wars have done what they will do. So the only thing left is how does the campaign allocate its resources. That’s where the ground game comes in.”
Q. Can you define what you mean by “ground game?”
A. The campaign whose party you seem to support wants you to vote, so campaign volunteers and surrogates focus on getting supporters with a low or middling record of turning out in prior elections to go to the polls. The ground game is direct contact by campaign surrogates, or the candidates themselves, with voters via face-to-face contact: meetings, knocking on doors, canvassing at shopping centers or other public places. The ground game also includes phone calls from volunteers or phone banks. In the digital age, voter mobilization includes other activities such as outreach via social media.
Q. How does that work?
A. Most of it involves how a campaign figures out whom to contact and how to communicate with them. Mobilization is a very important aspect of this, and there are two kinds. One is taking non-voters, or those unlikely to vote, and putting them in your stack. The other tool is persuasion, which essentially doubles a candidate’s vote because it involves taking someone out of the other party’s column. In principle, persuasion is a more effective strategy but in practice it’s difficult to convince the other candidate’s supporters to vote for your side.
- John William Burgess Professor of Political Science
- Professor, Political Science, Columbia University, 2011-2016
- Director, Institution for Social and Policy Studies, Yale University, 1996-2011
- A. Whitney Griswold Professor of Political Science, Yale University, 2001-2011
- Professor, Department of Political Science, Yale University, 1989-2001
Q. What does that look like this year?
A. In this election, roughly 90 percent of Democrats are for Hillary Clinton, and 90 percent of Republicans are for Donald Trump. The vast majority of registered voters will vote regardless of encouragement by campaigns. So the ground game is about persuading a relatively narrow but decisive slice of the electorate to vote. If you live in a battleground state, which in this election includes Ohio, North Carolina, Pennsylvania and Florida, chances are you will be hearing often from campaign volunteers on behalf of their candidate.
Q. Have we seen this in prior elections?
A. Not every election plays out the same way. Ironically, get-out-the-vote tactics played a fairly minor role in the very close election of 2000. It was Karl Rove’s insight that turnout would be the key to the 2004 election between John Kerry and George W. Bush. There weren’t many persuasive messages that would get people to switch their votes from one party to another, so it was all about getting a bigger army to the battlefield. In 2004, we saw an unprecedented surge in voter mobilization on the Republican side. There were armies of activists campaigning for the Bush/Cheney ticket. That was a lesson that the Obama campaign learned and utilized in 2008, when the Democrats had their own unprecedented surge in voter mobilization. Another factor in 2008 was that, while I wouldn’t say the McCain/Palin side was poorly organized, it was lacking in spirit. In 2008, I don’t think the Republican base liked McCain very much, or Romney in 2012.
Q. Is there a Trump voter mobilization effort in this cycle?
A. If Trump has a ground operation, it’s relatively small. The person who announced on day one that he has more than $8.7 billion hasn’t put money into organizing, whereas on the Clinton side there’s a sizeable effort—although perhaps not as big as Obama’s in ’08. The Democrats are outspending Republicans on TV by a margin of 7 to 1, and they’re out-organizing by maybe more than 7 to 1. We’ll know only what is working as we get into the final stages of the election.
Q. How has early voting changed the traditional get-out-the-vote effort?
A. Early voting changes the cadence of the ground operation. A surge of persuasive activity accompanies the initial distribution of mail ballots because roughly half of all recipients send them in immediately. Another surge accompanies the start of early in-person voting. After that, the focus switches to voter mobilization because a campaign can pare down its workload on Election Day if it can bring voters to the polls during the early voting period. Many registrars prepare a daily list of who voted, so that ground campaigns can focus their attention on those who have yet to vote.
Q. Where does the U.S. stand in terms of voter turnout?
A. Turnout in presidential elections has been climbing, which is a recent phenomenon. The election of 2000 was extraordinarily close—Bush won Florida by 537 votes— but turnout was relatively low, with only 107 million people voting. In 2004 that number was 124 million, and in 2008 it was 133 million. It dipped in 2012 to 126 million. On a percentage basis, only 59 percent of eligible voters went to the polls in the last presidential election. The U.S. has low turnout compared with other developed countries.
Q. What do you make of certain states that were reliably Republican in the past that are now tipping towards Democrats?
A. It really is stunning and an indication of how fast demographic change is proceeding. This year there are states in play that no one would have imagined a few cycles ago: North Carolina, which went for Obama in ’08 and narrowly for Romney in ’12, and Virginia, where decades of demographic change around the Washington suburbs has tilted the state’s political balance.
Q. You’ve been researching voter behavior for many years. Are we seeing anything different in this election cycle?
A. The two main differences are the asymmetry in media used by the campaign. Clinton is relying primarily on TV ads, while Trump is relying to an unprecedented degree on Twitter and on the free media coverage it generates. There’s also asymmetry in get-out-the-vote operations, which are extensive by Clinton, negligible by Trump. The get-out-the-vote operation is sometimes derided as the “field goal unit,” because it is unable to deliver more than a percentage point or two. In this election, however, a percentage point or two may be the difference between victory or defeat in key battleground states.
Q. A 2000 article by you and political scientist Alan Gerber suggested personal canvassing works better than other modes of contact. What is the best way to engage voters now?
A. That remains true. Personal contact—an authentic invitation to vote by someone who makes you feel that your voice matters—remains the most powerful method of voter mobilization.
Q. How does the internet fit into this? Promoted content from the candidates pop up in Twitter feeds and on Facebook every day—is that kind of advertising likely to sway voters?
A. That’s still unclear. Scholars have yet to assess in a rigorous manner the effects of Twitter on voting behavior. Social media is very important in organizing and getting people to come to rallies and give money, less important when it comes to persuasion.
Q. At Yale you taught a course called “Prejudice and Racial Intolerance,” which is relevant given this year’s election. What in this political climate makes it acceptable for candidates and their supporters to say things that would be disqualifying in prior elections?
A. As the parties become more ideologically polarized and as campaign strategies focus increasingly on mobilizing the base rather than winning over supporters of the opposing party, the level of invective had gradually increased. For many whites, especially men and especially conservatives from “red” states, Donald Trump represents resistance to ethnic change and a return to an earlier era in which whites felt no special need to cater to the concerns of non-whites. In some sense, this election is a test of whether one can form a winning coalition by appealing primarily to this segment of the electorate. In their post-mortem after 2012, the RNC declared that the party needed to attract a broader set of supporters, and the question, should the Republicans lose this election, is whether Donald Trump’s stance on immigration will prevent them from reviving their efforts to broaden the party’s base.
—Interviewed by Bridget O'Brian