US election monitor: When will we know the winner? Well, that depends

In this edition of the 2020 Election Monitor, we take a closer look at the timeline of the election night, the vote counting process and when we could know the election outcome. At the 2016 election, Donald Trump was called as the winning candidate by the network stations at around 8:40 CET. Trump had secured more than 270 electoral votes eight hours and forty minutes after the first state polls closed in Kentucky and Indiana (see full 2016 Election Night time line on page 3). With mail ballot being the big X-factor in the 2020 election, the election result could be delayed depending on how close the presidential race is.

We expect that we will know who the next president is on the morning of 4 November (CET) in the event of a blowout scenario, where the outcome can be decided solely on ‘fast counting' swing states such as Florida, Nevada, Texas, North Carolina and Arizona. This is because if Biden is to win all the leading/solid Biden states on the RCP electoral map, he will need only 38 of the toss-up votes to secure a majority. The presidency is thus as good as secured for Biden if he wins Florida and the appurtenant 29 electoral votes. According to FiveThirtyEight, Trump has a <1% chance if he loses Florida. In Florida, election officials say there's a good chance that they will have the overwhelming majority of ballots counted by midnight, when we will also most likely know if Biden has a blowout victory.

In a tight race scenario, we could experience a similar situation as in 2000 with recounting, where Al Gore conceded over a month after Election Day. Generally, the time and potential delay of the announcement of the winner is much more uncertain with a close race. The remaining considerations are mostly relevant in the case of a tight race. Remember that all election disputes must be resolved no later than 8 December, according to US law.

To account for the large number of mail-in votes, most states have allowed for early vote processing (signature confirmation, photo-ID, etc.). However, both Wisconsin and Pennsylvania only allow the process to start on Election Day – see Chart 4 overleaf for an overview. The "Rust Belt" states are decisive swing states, where Trump has >50% chance of winning the overall presidency if he claims at least one of the three states (Michigan included), according to FiveThirtyEight. Knowing the outcome of these states could be the tipping point that secures either candidate the majority. Additional to the late processing, neither Wisconsin nor Pennsylvania have robust histories with mass mail-in voting, which further adds to the probability of a delayed result. Full details on vote counting legislation in key swing states as well as the likelihood of late results in each state are shown in Chart 7.

Apart from the slow counting in the key swing states, the main reason for a delayed announcement lies in the fact that almost half of the states merely require mail-in ballots to be postmarked, rather than received, by Election Day (see Chart 1 above). This includes some already-mentioned key swing states: Pennsylvania and North Carolina, but also swing states where candidates are currently neck-to-neck, such as Ohio. Trump leads the RCP Ohio poll by 0.6pp in the most recent poll.

Let us take a look at who is currently leading the presidential race with 4 days remaining until the election. Biden is leading the national polls by 51% to Trump's 43.5%, the gap between the two candidates being more than twice as large as this point in time in 2016. Prediction markets put the likelihood of a Biden victory at 60%. The blowout victory scenario triggered by a Biden win in Florida looks like a 50-50, as Biden and Trump are now neck-to-neck in Florida after a big Trump rebound.

Without Florida, Biden's main victory path is by winning the "rust belt", where he is ahead in all three states (3.8pp in Pennsylvania, 8.6pp in Michigan and 6.4pp in Wisconsin). If Biden wins the Rust Belt states he would need only nine more electoral votes to win the majority, making Florida redundant. In this scenario, Biden only needs to win Arizona, where he is currently leading by 2.2pp.

Another factor that may favour a Democratic victory is that voter turnout is on track to be the highest in centuries according to the news sites CNN and The Guardian. The US Elections Project predicts a record turnout of about 150 million, representing 65% of the eligible voters, versus 139 million or 55.5% in 2016. More than a third of registered voters -- over 75 million Americans or about 36% -- had already voted six days before Election Day. According to Politico, young voters are one group that especially drives up the turnout. In Florida, young voters have already cast 3.5 times more early votes than in 2016, a trend that repeats itself across all states. According to a Harvard Youth poll, people under the age of 30 favour Biden over Trump by 65–25.

The Republicans are closing in on the Democrats in the Senate election with prediction markets giving them a 44% chance of winning the majority. The Democrats have lost momentum after Arizona went from leaning Democratic to becoming a toss-up state. Republicans need to win only 5 of the 9 toss-up states to secure a majority in the Senate. The Republicans currently lead the state polls in 3 of the 9 toss-up states.

Overall, the state polls and the overall probability favour Biden. But as we discussed in last week's edition, if we were to experience a similar difference between the polls and the actual election results as in 2016, it would mean that Trump would win the 2020 election. Ranking the election outcomes based on likelihood, we believe that the most likely scenario is 1) Democratic clean sweep 2) Biden with a divided congress 3) Trump with a divided congress 4) Republican clean sweep.

Download The Full US Election Monitor