Intense preparation helped Hillary Clinton command the podium in her first debate with Donald Trump in September 2016—the most-watched presidential debate in US history.
In one of the debate’s most memorable moments, Clinton called out Trump for ridiculing a beauty pageant winner. Trump, admittedly less prepared for the debate, struggled to respond. At one point, Clinton, reveling in her success, shimmied with meme-worthy glee.
Every major polling outfit declared Clinton the debate’s victor the next day. But it didn’t make a difference: Trump went on to win the election. That’s because debates have only a negligible effect on voters’ candidate choice, according to new research from Harvard Business School. In fact, 72 percent of voters make up their minds more than two months before the election, often before candidates square off. And those who shift to a different candidate closer to the election don’t do it following TV debates.
"We find that debates don’t have any effect on any group of voters."
“There’s this perception that debates are this great democratic tool, where voters can find out what candidates stand for and how good they really are,” says Vincent Pons, assistant professor of business administration. “But we find that debates don’t have any effect on any group of voters.”
Pons and co-author Caroline Le Pennec-Caldichoury, a doctoral candidate in economics at the University of California, Berkeley, mined pre- and post-election voter surveys from 61 elections in nine countries, including the US, UK, Germany, and Canada.
Among the 172,000 respondents in the study, almost 80 percent had watched a debate. The team compared vote intentions in the pre-election survey to actual votes reported in the post-election survey. That allowed the team to pinpoint exactly when a voter decided on a candidate—before or after a debate—without having to rely on voters’ memories. They found:
- A large fraction of voters change their minds during the campaign. The percentage of voters who had settled on a candidate rose by 15 percentage points, from 72 percent to 87 percent, during the two months before an election, indicating that the information received in this period matters overall.
- Voters who switched candidates didn’t change their policy preferences. Voters’ policy views remained consistent, despite the onslaught of campaign ads and social media posts. Those who voted for a candidate other than the one they originally planned did so only when new information emerged about the candidates or new issues gained salience during the campaign.
- Debates didn’t shift elections’ outcomes. High-profile TV debates didn’t increase the fraction of voters who had settled on a candidate nor the candidates’ vote shares. This suggests that voters who shift to another candidate are influenced by other types of information, such as media coverage of the campaign or personal discussions.
- Debates failed to persuade even young and less educated voters. One might expect debates to appeal more to young and less educated voters, who have less preexisting information, whether they directly watch debates or hear about them from friends or media coverage. But they don’t.
“It is difficult for candidates to change people’s minds, and this does not happen on TV or the radio,” Pons and Le Pennec-Caldichoury write in a working paper released last month, Vote Choice Formation and the Minimal Effects of TV Debates: Evidence from 61 Elections in 9 OECD Countries
The business of TV debates
Televised debates have been an election mainstay since 1960, when John F. Kennedy’s youthful vigor overshadowed Richard Nixon’s flat pallor on American screens. Debates help candidates reach millions of voters in a single event, so candidates often prepare rigorously, hoping to gain an edge with undecided voters. And major TV networks usually line up pundits to provide hours of analysis afterward.
“These debates are often great moments of television,” Pons says. “We all remember the gaffes that candidates make as well as the memorable comebacks. Debates are entertaining.”
Who could forget Trump calling Clinton a “nasty woman” during their final debate in October 2016? Or Ronald Reagan’s witty response to concerns about his age that arose in his 1984 debate with Walter Mondale: “I am not going to exploit, for political purposes, my opponent’s youth and inexperience.”
The opportunity to speak directly to voters has proved alluring enough that a growing number of countries, from Sweden to New Zealand, now hold debates, often with much media fanfare.
The best way to win hearts and minds
Democratic US presidential candidates took to the debate stage for the fifth time this week. While Pons didn’t study the role of debates in primary campaigns, memorable moments from earlier debates this year have helped candidates inspire donors and gain media attention.
While one of these candidates might square off with Trump at the podiums next year, the election’s winner will ultimately depend on a campaign’s ability to reach voters in person. Even in a high tech world, the best place to change the minds of voters, especially disenchanted voters, is on their doorsteps, Pons says.
Voters who dismiss the canned responses of a well-groomed candidate might listen to an activist from the same town or a similar background. And the fact that these canvassers tend to be unpaid volunteers makes these conversations even more meaningful.
“The discussions that voters and canvassers have on doorsteps are very personal,” says Pons, whose past research found that even a five-minute conversation can convince a voter to support a different candidate. “They’re a very effective way to mobilize nonvoters and persuade those who are undecided.”
Getting the ground game right
Candidates who accept the limited benefits of TV and shift resources to in-person outreach—whether by the candidate, a proxy, or a volunteer—might improve their odds of winning, Pons says. But that’s not how campaigns spend most of their time and resources. Among the voters Pons and Le Pennec-Caldichoury studied, less than half had some form of contact with any candidate’s campaign, and only 28 percent had received a visit.
“This fraction only increases late in the electoral season, when most voters have already made up their mind,” the team writes.
Given the importance of a campaign’s ground game, Pons says that governments should consider ways to improve the quality of information that voters receive. That might involve setting campaign finance regulations that level the playing field and give candidates equal access to voters, and monitoring the accuracy of handouts and social media posts that campaigns share.
“It’s important to make sure that all campaigns are able to get their message out through these powerful channels,” Pons says.
About the Author
Danielle Kost is senior editor of Harvard Business School Working Knowledge.
[Image: David Goldman/AP]
Related Reading
- Voter ID Laws Don't Work (But They Don't Hurt Anything, Either)
- Why People Don’t Vote--and How a Good Ground Game Helps
- 'I Know Why You Voted for Trump' and Other Motivation Misperceptions