Whether you look at nationwide polls or surveys of likely swing states, Joe Biden is comfortably ahead of Donald Trump.
US polling guru Nate Silver’s FiveThirtyEight model currently gives Biden an 85% chance of winning.
The likelihood of Trump turning things around looks minimal as there seem to be relatively few voters who say they are persuadable.
One recent CNN poll found that just 8% of likely voters said they might change their mind, while only 1% had no preference between the candidates.
Despite all the evidence that Trump is on course to be a one-term President, my perception is that few feel like they can say with any level of confidence that Biden will win.
Part of this will be thanks to the fact that Trump outperformed his polling to win in 2016.
People are understandably suspicious of believing Biden is a sure thing when they thought the same of Hillary Clinton four years ago.
And there have been plenty of other recent examples of electoral volatility to make people shy of making predictions.
But is there more to it? Is there any evidence to substantiate scepticism about the headline polling numbers this time around?
For Trump supporters looking for a glimmer of hope or Biden fans wanting to have their expectations lowered, there are some data points to cling on to.
First, the economy has returned to be the most salient issue. Trump has positive approval ratings on the economy and, in most polls, is ahead of Biden on the issue.
Trump is also ahead on “strong and decisive leadership”. This was also the case in 2016 versus. Clinton. Perceptions of strong and decisive leadership is a disproportionately important factor in voter preference.
For example, we saw in the UK’s 2019 general election that people don’t need to “trust” a leader in order to overwhelmingly elect them to run the country.
Another cause for concern for Team Biden should be that the American people predict that Trump is going to win.
The public’s opinion on the likely victor has been an accurate predictor of the winner of the popular vote (though not the winner of the electoral college, which decides the result) since the question started being asked in 1996.
Another data point that should give cause for concern to Biden’s team is his earned media share.
News media and content shared organically on social media provides the information most people use to educate themselves on how to vote, so earned media share is often a good predictor of electoral success.
While there has been lots of hearsay about Biden avoiding the press and Trump once again dominating the headlines in the build-up to a presidential election, hard data on earned media share for this cycle is hard to find.
However, one recent survey asked people whether they had “Heard about [Trump/Biden] in the news”; 65% of people said they had heard about Trump “often” in the news and only 27% said the same for Biden.
And we can clearly see who is winning the earned media war on Facebook, the most important social media channel at election time, thanks to its reach.
Trump has more page likes on Facebook, gets more engagements per post and his content gets organically spread further; Biden averages 130,737 views per video post compared with Trump’s 441,722 views per video post.
And while Biden and his supporters have outspent Trump and his supporters on advertising – the Democratic candidate alone has spent more than half a billion dollars on advertising so far – that is not a guarantee of success.
Biden spent less on advertising in Super Tuesday states than all of his main competitors in the Democratic Party primary election and yet he won a majority of states on that all-important day.
And let’s not forget that in 2016 Hillary Clinton spent nearly three times more than Trump on advertising.
Despite Biden spending more money on advertising, people seem to be remembering his ads less than Trump’s.
And it’s a similar story for online ads with 26% of respondents having seen “many” for Trump, against 14% having seen “many” for Biden.
Biden’s advertising has lacked message consistency and has a tendency to slip into vague platitudes around “unity”, so it’s believable that it’s not cutting through.
“Share of internet search” has been shown to be a good leading indicator of market share in the commercial world and there is evidence that online search volumes are a predictor of success in US Presidential elections too.
Trump is getting significantly higher volumes of people searching for him online than his rival.
And the final aspect of this election that should make even the most confident of forecasters sceptical is the context of the Covid-19 pandemic.
We’ve never had an election in a situation quite like this one and the impact of things like a dramatic increase in mail-in ballot papers and a population feeling deeply uncertain about the future are impossible to predict.
With all that said, there are far more data points which imply a comfortable Biden win.
But if in three weeks' time, in the aftermath of another Trump upset, we end up looking back for signs that could have told us of what was to come: they do exist.
Benedict Pringle is founder of politicaladvertising.co.uk
Photos: Chip Somodevilla/Getty Images (Biden), Win McNamee/Getty Images (Trump)