Kamala Harris has a narrow lead over Donald Trump in the polls, but it may not be enough to win the 2024 US presidential election, according to a poll of polls by The Times and The Sunday Times.
With less than two weeks until Americans vote, Harris, the Democratic nominee, is polling ahead of Trump, the Republican candidate. However, she may not have enough votes in the crucial swing states that will decide the election in the electoral college.
Between now and November 5, we will be publishing a rolling average of opinion polls from more than a dozen pollsters.
It has been a lively campaign. Yet for most of it – despite two assassination attempts on Donald Trump, and two eventful presidential TV debates – the polls have been fairly static.
However, in the past few weeks, there has been some narrowing of the polls.
Can Kamala Harris win the US election?
Harris, the vice-president, and her running-mate Tim Walz, the governor of Minnesota, initially enjoyed an upbeat start to their pitch for the White House.
But while she may be more popular than Biden, popularity alone does not always win presidencies. The Republicans have not won the popular vote since 2004, while in 2016 Hillary Clinton won more votes but lost the election.
• Who could serve in Harris’s cabinet?
Because of the electoral college system, in which the winner of each state takes all of its electoral votes, the result will almost certainly come down to a small group of people in the seven closest states. In 2020 Trump won one of them, North Carolina, while Biden won the rest.
Who will win the swing states in 2024?
If we take an average of every statewide poll conducted in the past three weeks, Harris is ahead in two of these seven: Wisconsin and Michigan, both in the rust belt.
However, a small polling error, or a shift in opinion by a few thousand voters, could tip them the other way.
Trump, meanwhile, appears to be ahead in Georgia and North Carolina, both with 16 electoral votes each. Crucially, he may also be ahead in Pennsylvania, the biggest prize with 19 electoral votes. No Democrat has won the White House without winning Pennsylvania since 1948.
State-level polling is incredibly difficult. Margins of error tend to be large: in reality, any of these seven states could easily be leaning the other way. There are signs that Harris is doing better in these states than Biden was, but it remains to be seen whether it will be enough.
If an election were held today, and these states voted as the polling averages suggest, then assuming all other more partisan states voted as expected, Trump would win the election with 287 electoral college votes.
However, if Harris performed just a fraction of a percentage point better than expected in Pennsylvania, she would win. The presidency remains on a knife-edge.
• What will happen if Trump wins?
Who is voting for Kamala Harris?
Support for Harris follows a similar pattern to support for Biden in 2020: she is more popular among the young, the college-educated and black voters.
Yet polling conducted by the Pew Research Centre before and after Biden pulled out of the race suggests that Trump has not actually lost that many voters to Harris. Rather, Harris appears to have won round many who were going to back the independent candidate Robert F Kennedy, who has since suspended his presidential campaign and endorsed Trump.
Harris’s poll ratings are 15 points higher than Biden’s among 18 to 29-year-olds, and 21 points higher among Hispanic voters.
However, there are some signs that Donald Trump is doing much better among Black and Hispanic voters than in 2020.
• What will happen if Harris wins?
Can we trust the polls?
On the eve of the 2016 election one polling model gave Hillary Clinton a 98 per cent chance of winning the presidency. Eight years on, that mistake still stalks the polling industry. “Polling has struggled when Trump is on the ballot paper,” says Dr Mark Pack, author of Polling Unpacked.
In 2020, polls had Biden about 7.9 points ahead; in reality his lead was just 4.5 points. This time, given how few votes separate the final outcome, a similar error could see pollsters fail to predict another Trump victory.
There is a perception, perhaps fuelled by the recent performance of pollsters, that bookmakers have access to better data than polling companies themselves. This is largely false, but we can nonetheless infer a kind of betting average from the latest bookmakers’ odds.
On July 16, less than a week before Biden stepped down, Trump had a 66 per cent chance of winning the presidency, according to a rolling average by RealClearPolling, with Biden’s chances at just 19 per cent. Yet since early August bookmakers have had the two neck and neck.
• Pick your path to the White House
Methodology
Instead of using complex models, we simply take a rolling average of every survey published by mainstream polling companies. All polls, regardless of who did them, are treated equally, but we consider only the very latest poll from each organisation. If a poll includes figures for both registered and likely voters, we opt for the latter. For both the national averages and state-level averages, polls older than 21 days are excluded.
This article was first published in January 2024 and is updated weekly.
