Kamala Harris v Donald Trump. The one question that everybody wants answered ahead of this election is ‘Who is going to win?’.
One of the best ways to predict who will be the 47th President of the United States is to look at the polling numbers. Registered voters are asked how they intend to vote throughout the campaign and these numbers are crunched and tracked by thousands of analysts worldwide to predict the outcome of the election. You might think that the best way to find out who will win the election is to measure who will win the most votes, but this isn’t always the case.
Harris has held a consistent lead over Donald Trump in the popular vote since she accepted the Democratic candidacy in the summer. On 1 October, she was averaging 48.6% of the vote to Donald Trump’s 45.9%, but that gap has narrowed as election day has neared. Most national polls still have Harris in the lead, but with only a 1% advantage, well within the margin of error.
However, due to the electoral college system in place in the United States, the placement of these votes is often far more important than the total. Each state has a certain number of ‘college votes’, decided by its population size, to give to whichever candidate receives the most votes from its electorate.
Considering that most states have voted the same way for many years, the candidates will have an assumed number of votes in the bag. For example, Texas has voted Republican in every election since 1980, and Washington has voted Democrat in every election since 1988. The states that alternate back and forth are called ‘swing states’. These are Arizona, Georgia, Michigan, Nevada, North Carolina, Pennsylvania, and Wisconsin.
In the final polling completed by Redfield and Wilton between 28th-31st October, they found Trump leading Harris in four out of the seven swing states. However, in every state polled, they were either tied or the lead was within the margin of error. In the last election, Biden won Arizona by only 10,457 votes and Georgia by 11,779. These are a tiny proportion of the vote share (0.3% and 0.82% respectively) in those states and with many states still far too close to call, there could be a similar outcome tonight.
Polling numbers also differ wildly between different demographic groups. According to a USA TODAY/Suffolk University poll in October, Trump lead among men by 53% to 37%, while conversely Harris lead with women by 53% to 36%.
There are similar discrepancies among ethnic and religious groups. Over three-quarters of black voters say that they would favour Harris over Trump nationally. Although Black people make up only 13% of the electorate nationally, they are a much bigger proportion of voters in some of the swing states Harris will be hoping to win.
Whilst Harris is also ahead of Trump with Latino voters, the former President has improved his averages with them from his defeat to Joe Biden. Harris also appears to be struggling with Muslim voters over her stance on the conflict in the Middle East. In fact, according to a November poll by the Council on American Islamic Relations, Harris was second behind Green Party candidate Jill Stein among 1,449 Muslim voters.
With voter bases so polarised and poll numbers indicating a dead-heat, the election could be decided by turnout. Trump has undermined the postal voting service with claims of electoral misconduct, and this may encourage Republicans to vote in-person, when they are more likely to be disrupted by long queues or a busy day. It is also not clear whether Harris has done enough to enthuse voters or provided enough information about her agenda and her political identity outside of the Biden administration.
Polls have painted a picture of a nation divided and the race for the White House on a knife’s edge, but how do we know we can trust them? The short answer is, we don’t. In both 2016 and 2020 pollsters substantially underestimated Donald Trump’s total. Producing accurate polling figures in an increasingly divided and polarised America is difficult – especially when methods of communication differ so much between demographic groups. But the pollsters have made several changes to their methods since 2020 and had a particularly good accuracy rating during the 2022 midterms.
Others have suggested that tracking the betting markets is a better way to predict the outcome. They favoured Harris when she accepted the nomination, before backing Trump in mid-October, before finally shortening odds on a Harris victory in the final days of the campaign.
Nobody is certain on who will win the election. The pollsters, the bookies – and, it seems, the electorate – have flip-flopped between Harris and Trump for the last four months. Ultimately, it will be the people, not the pollsters, who decide this election, and ‘the people’ are proving increasingly unpredictable.