Hillary Clinton versus Donald Trump: do presidential debates influence the election result?

As the US presidential candidates prepare for their first debate, can such events affect the outcome? We look at past debates and the polls that followed.

Sign Up

Get the New Statesman's Morning Call email.

Presidential debates are big events. 67 million Americans tuned in to the first encounter between Barack Obama and Mitt Romney four years ago. The candidates spend a lot of time preparing for their bouts, and reaction to them dominates election coverage for days afterwards.

But all this attention on the debates exaggerates their importance when it comes to determining the next occupant of the Oval Office. That’s not to say that they have no effect – indeed, there are a couple of elections that could easily have turned out differently without them, as we’ll see below – but debates are rarely the “game changers” they’re often billed as.

According to political folklore, it was Richard Nixon’s sweaty face and cheap make-up in the first ever televised debate that handed the 1960 election to John F Kennedy. It’s certainly possible that the debates helped Kennedy to his narrow victory that year, but there isn’t much hard evidence. Polls by Gallup showed Kennedy going from 1 percentage point behind Nixon before the debate to 3 points ahead afterwards – but that could be down just as easily to sampling error as to sweatiness.

The other result that may have hinged on the debates was, unsurprisingly, another very close election: the 2000 race between Al Gore and George W Bush. Gore entered the debate with a 1.8-point lead, according to Nate Silver’s averages of the polls taken the week before the debate.

But Gore was widely criticised for sighing while Bush was talking and, in the average of polls taken the week after the debate, Gore now trailed by 1.5 points. He did recover to win the popular vote by half a point but, of course, not the presidency.

Challengers to incumbent presidents have tended to see their poll numbers rise after the first debate, perhaps because it’s a chance to make their pitch to voters who don’t know much about them. The largest example of this is the most recent: when Romney got the better of Obama in their first debate four years ago. Obama’s poll lead shrunk from 5 percentage points on the eve of that debate to less than 1 point a week later.

However, as with convention bounces, any poll bounce a candidate receives from a debate is likely to be short-lived. In the end, Obama won the 2012 election by 3.9 points – almost exactly the margin FiveThirtyEight was forecasting just before his poor first debate.

So, in general, debates have less of an effect on the outcome than the breathless coverage of them might suggest. But this is a close election – Hillary Clinton currently leads Donald Trump by just 2 points. Even a small debate bump for one of them could change the complexion of the race significantly. And there are some good reasons to think that the debates could have a bigger impact this year.

In the past, debates have tended to pit experienced, competent, well-prepared debaters against each other. This time, Clinton – who’s performed well in lots of debates against Obama in 2008 and Bernie Sanders this year – faces Trump, who’s never debated anyone one-on-one and didn’t do terribly well in debates with his competitors for the Republican nomination.

Clinton’s been preparing assiduously: studying briefing books and clips of Trump, and taking full days out of her campaign to practise with her experienced team. Trump, meanwhile, has been “testing out zingers” with Rudy Giuliani and Roger Ailes over cheeseburgers at his golf course.

However, it’s not the candidates’ performances that matter, so much as what the media says about them. In 2004, Kim Fridkin and others found that viewers who watched NBC’s relatively pro-Bush coverage of the final debate were more likely to declare him the winner than those who just watched the debate, or who watched it with CNN’s online commentary.

And as we’ve seen plenty of times already in this campaign, much of the media holds Trump to a lower standard than Clinton. Don’t be surprised, therefore, to hear positive reviews of Trump “beating expectations”, even if he doesn’t perform as well as his opponent.

There is also an unusually large number of voters who are not currently planning to vote for either major candidate. Right now, about 18 per cent of voters are either undecided, or say they’ll vote for one of the “third-party” candidates, such as Libertarian Gary Johnson or Green Jill Stein. That compares with roughly 8 per cent at this point in the 2008 and 2012 elections; in each case the eventual third-party vote totalled less than 2 per cent.

That means there’s a bigger than usual pool of potential voters for the candidates to tap into with a strong debate performance, especially with neither Johnson nor Stein on the stage.

So the debates might well make a difference to this campaign – especially as it’s so close right now. But don’t expect a “game changer” where one candidate seals victory with a clever zinger or witty put-down. It takes more than that to become President.

Jonathan Jones writes for the New Statesman on American politics.