We’ve all been there. Before a big event, we plan what we’ll do and say, rehearsing every detail. But once we’re there, it’s different. The atmosphere, sense of responsibility, and the people around us make things unlike our solitary rehearsals.
Sometimes, that big event is an election. Each of us who changes our minds in the final days could influence the results, making them subtly—or drastically—different from what polls predict.
This shift in voter opinion isn’t unique. It’s a common pattern. Instead of taking polls as gospel, it’s worth exploring why people change their minds and what polls miss, to understand who might win or lose in that final week before election day.
According to our research, between 20 and 30% of voters make up their minds or change them in the final week of a major election, half of them (10-15%) on election day itself. Some might choose to vote for a candidate instead of abstaining. Others might switch their vote or part of it (like in the U.S., where a single ballot can include multiple votes). This is true in highly contentious elections (polling for the 2016 Brexit referendum in the UK confirmed these proportions), and it can be even higher in less significant elections (data from a 2012 referendum on children’s rights in Ireland showed that a large majority of voters were last-minute deciders or switchers).
It can be hard to see how much individual change happens because many voters cancel each other out. For example, if someone intended to vote for Democratic candidate but decides to abstain, while another planned to abstain but ends up voting for Harris, that’s two people changing their minds, but the overall result doesn’t reflect it. In other cases, changes go in the same direction and lead to unexpected electoral outcomes, like in the [Name of Election] in July, where the National Rally came in a distant third despite polls predicting they would win a majority (sometimes an absolute one).
Crucially, not all citizens view their role as voters in the same way. Comparing an election to the Superbowl, some voters see themselves as “supporters” who will likely vote for their “camp” regardless, while others see themselves as “referees” who will assess the candidates and programs and try to pick the best person for the country.
As a result, while many focus on “polarization,” our research suggests that an increasing proportion of people—especially young people—are hostile rather than polarized. They are critical of everyone, not just those who vote differently.
In other words, if the U.S. electorate were merely polarized, people would vote enthusiastically for either Kamala Harris or [Name of Candidate] and flock to the polls. But instead, many will abstain or vote for third-party candidates. Others who go to vote won’t do so because they like either candidate but because they dislike one or the other intensely. This is a 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 with strong criticism of both, what tips the balance in the end is hard to predict.
In many ways, next week’s decision will not be made by hardcore supporters for either camp but by many disillusioned citizens across the U.S. who are 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 prefer Republicans or Democrats, or the policies they want for the country. Increasingly, people want to use democracy to claim that the system isn’t working or to feel respected and listened to by political elites, not to influence policy or seek representation.
This sense of negativity comes at a cost for a democracy. When a significant proportion of voters will likely vote “against” something rather than for it, it creates frustration among citizens and often towards other citizens rather than politicians and elites. Over the years, politicians have become both worse losers and worse winners than they used to be, and this makes it harder for elections to bring the sense of resolution society needs 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.