US 2020 election update: Both a Trump win and Biden landslide are within the margin of error

On the eve of the first presidential debate, nine states are within a three-point margin of error for the leading candidate.

Sign Up

Get the New Statesman's Morning Call email.

With less than six weeks to go until the US presidential election, the New Statesman is launching a weekly series of data-based insights, rounding up what you need to know about the race for the White House.

How close is this election?

The campaign has so far seen Democratic candidate Joe Biden maintain a national lead over Republican incumbent Donald Trump of eight points – with little sign of it abating.

In the last 20 years, we have not seen a lead as steady and sturdy as Biden's. Though Trump is performing marginally better than in 2016, Biden – owing to the poor position of third party candidates this time around – is in a much stronger position than Hillary Clinton was at this point in the race.

In the key swing states, however, things are not clear cut.

The New Statesman’s US election model finds, on the eve of the first presidential debate, that nine states are within the three-point margin of error for the leading candidate.

Donald Trump is an error margin away from winning re-election
Data from the New Statesman's US election model, taken Sept 28

Florida, Georgia, North Carolina and many of the Rust Belt states give leads for Donald Trump or Joe Biden that are narrower than the typical margin of error one needs to account for when modelling with polling.

Texas and Arizona are also tight. The projected two-party share for Texas has Trump with 51.4 per cent to Biden with 48.6 per cent. Arizona, meanwhile, has Biden on 51.7 per cent and Trump on 48.3 per cent.

While our model gives Trump less than a one in five chance of winning re-election, he is still within the error margin of victory.

If all the states with the candidate ahead by 3pts or less were won by Trump, he'd squeak through and secure a second term. If all were won by Biden, we'd be looking at a landslide victory for the Democratic challenger.

Trump will take the stage on Tuesday night with his odds of winning having dwindled since mid-September in much of the Midwest. In Florida, however, owing to growing voter uncertainty among the Hispanic population, his chances have considerably improved.

Will the debates throw him a lifeline? In 2016, after the third presidential debate, Trump’s odds at winning surged, according to the FiveThirtyEight model. His probability of winning jumped from 13 per cent in mid-October to 29 per cent on election day.

This year, we go into the debating season with a seemingly more stable electorate than anything we’ve seen for quite some time – but also one that is at risk of being swayed by certain contentious issues.

The SCOTUS vacancy

The death of Supreme Court Justice Ruth Bader Ginsburg has opened up an opportunity for Trump and the Republicans. Were they able to rush and fill the vacancy before the November election, it would create the most conservative Supreme Court in 70 years. 

But such a move could prove unpopular. Most voters agree with Biden that the Supreme Court vacancy should be filled after the November election, and polls conducted since Ginsburg’s death show, in most cases, large majorities of the American electorate are in favour of waiting. In some key swing states, such as Michigan and Wisconsin, majorities of 54 and 56 per cent of voters respectively are in favour of the view that “the result of the election should decide who fills it”. 

The data here bodes well for Bidenland – but the move to replace Ginsburg is unlikely to matter as much as some may wish. Those voters already enthusiastic about the SCOTUS vacancy are, really, just that: enthusiastic voters. Voting intentions already factor them in. Those that value SCOTUS and matters of constitutional import are not swing voters, and are unlikely to have had their vote altered as a consequence of the events.

While the SCOTUS vacancy might not change votes, it could, however, dampen the enthusiasm to go to the polls. Voters registered as independents are, in most polls, breaking for Trump by the narrowest of margins. In 2016, they made up more than one-third of the voting population, and broke 46-42 for Trump. Little has changed since, but the SCOTUS vacancy, as with other issues, appears to be chipping away at confidence among registered independents in their vote for the Republican incumbent. The latest survey from YouGov reports the death of Ginsburg makes registered independents 34 per cent more likely to vote Democratic, and 17 per cent more likely to vote Republican.

Will Trump's tax returns make the difference?

On Sunday, the New York Times reported that in his first year at the White House Trump paid $750 in federal income taxes. Knowing that they pay more in taxes than the president will likely irk many voters, but the reality is this is a news story that is unlikely to do much in the way of changing votes. As our US editor Emily Tamkin writes, Trump is, after all, no stranger to scandals or exposés. 

There is little favourability towards Trump among those unsure about their vote, or those sitting the election out. Trump’s Marmite tendency means those already committed to him are doing so with great enthusiasm, and little room for waverers. In 2016, Trump’s supporters were more enthusiastic than Clinton’s, and the same can be said, albeit to a much lesser extent, in 2020.

While the tax returns revelation may not be what makes voters abandon Trump, it risks becoming a talking point in an already rocky campaign for the president.

In 2016, the same might have been said for Clinton. In the closing weeks of the campaign, the mood music was dominated by the so-called email scandal. Again, the issue may not have changed many votes, but it did motivate certain voters and demotivate others.

We don't yet know how, or if, the Trump tax story will shape the weeks ahead, but it risks damaging enthusiasm to the president's vital margins.

 Ben Walker is a data journalist at the New Statesman

Free trial CSS