What Polls Won’t Tell You About Voting Patterns

We have all been there before. Before a big momentous event, we think of what we’ll do, what we’ll say, we have it rehearsed to the last detail, but once there—Well, once there, it’s not quite the same. Somehow, our little rehearsals had missed out on the atmosphere, the sense of responsibility, the people around us who make things quite different from what we had imagined alone.

Sometimes, that big momentous event is an election, and every one of us who ends up making up (or changing) our minds in the final days of the campaign could be responsible for the results being subtly—and sometimes not so subtly—different from what polls suggest.

Despite what pundits say, the 2024 U.S. election is not unique. It’s all part of a bigger pattern. So instead of taking the polls at fact-value, it’s actually worth exploring why people change their mind and what the polls can’t tell us to inform us who may be on the winning or losing end of that final week before election day.

According to our research, in a typical major election, between 20 and 30% of voters end up making up or changing their minds in the final week of the campaign, half of them (so 10-15%) on election day itself. For some people, this will mean choosing to vote for one of the candidates instead of abstaining. Or the other way round, for some it will mean actually switching their vote or part thereof (in the case of the United States where a single ballot can include many different votes). This is true of highly contentious elections (for instance, polling by Lord Ashcroft confirmed those proportions in the highly divisive 2016 Brexit referendum in the UK), and those proportions can be even higher in less salient elections (data on a 2012 referendum on children rights in Ireland showed a large majority of voters were in fact last week deciders or switchers).

At times, it may be hard to realize the prevalence of individual change because a lot of voters will cancel each other out. For instance, in the case of the U.S. election, if one person intended to vote for Democratic candidate Kamala Harris but decides to abstain, whilst another planned to abstain but ended up voting for Harris, that is two people changing their minds, but there will be no visible effect when you look at the overall result. In other cases, those changes tend to largely go in the same direction and result in electoral outcomes no poll had foreseen, as was the case with the French elections in July, which saw the National Rally come a distant third whilst every single poll had predicted that they would at least win a relative majority (and sometimes an absolute one).

Crucially, not all citizens see their role as voters in the same way. In fact, comparing an election to the Superbowl, we find that some voters see themselves as “supporters” who will likely vote for their “camp” whatever may be, but others see themselves as “referees” who will assess the worth of the candidates and their programs, and try to pick whoever they think would be best for the country.

Read more: Why You Shouldn’t Let Election Polls Stress You Out

As a consequence, whilst many eyes are on “polarization,” our research suggests that an increasing proportion of people—particularly among young people—are hostile rather than polarized. Meaning they tend to be quite critical of everyone rather than only those who vote differently from them.

In other words, if the U.S. electorate was merely polarized, people would be voting with enthusiasm for either Kamala Harris or Donald Trump, and would flock to the polling stations en-masse. But instead, many people will abstain or vote for third-party candidates. Others who go to vote will not do so because they like either candidate but because they will dislike one or the other of them very much. This is the symptom of an increasingly hostile—rather than polarized—society, and the effects of hostility in those final days are harder to assess. Many people say they want to vote for the “lesser of two evils,” but because they have strong criticism to address to both, what tips this very negative balance in the end is very hard to tell.

In many ways, next week’s decision will not be made by hardcore supporters for either camp but rather by the many citizens across the U.S. who have become largely disillusioned— sometimes hopeless about democracy and society and will use their ballot to express a wide range of fears and frustrations. Much of this will have nothing to do with whether they like Republicans or Democrats better, or the public policy they would prefer for the country. More and more, people want to use democracy to claim that the system is not working or feel respected and listened to by political elites than to influence policy or seek representation.

This sense of negativity has a cost for a democracy. When significant proportions of voters will likely vote “against” something rather than for it, this creates frustration amongst citizens and often towards the citizens themselves rather than politicians and elites. Over the years, politicians have become both worse losers and worse winners than used to be the case, and this makes it harder and harder for elections to bring the sense of resolutions society needs to be able to breathe, move on, function smoothly, and open the door for new discussions and healthy disagreement.

If there is one thing polls tell us for sure, it is that a significant proportion of Americans are, at any rate, worried about the day after, and unconvinced that it will resolve any of the rifts society has suffered from across the years.