by Holger Schmieding*
With luck, we will know the result of the U.S. elections exactly two weeks from now, on the morning of November 4th.
No more a mainly COVID 19 campaign
As it stands, Joe Biden remains the favorite to win. But on two of the three key metrics (betting odds and opinion polls), his lead over Donald Trump has narrowed somewhat.
After Trump’s return to the campaign trail following a brief COVID 19 illness, the focus of the campaign is less on his mishandling of the pandemic than it briefly was some nine days ago.
2016 all over again?
The experience of 2016 with a late rise in support for Trump in some key states followed by his upset victory over Hillary Clinton also suggests that the race remains open.
• According to the bets covered by electionbettingsodds.com, the probability of a Biden win has declined to 60.5% from 65.3% on October 12th.
Conversely, the probability of a Trump win now stands at 36.6% versus 32.1% on 12 October. At the same stage of the 2016 election, betting markets had seen Hillary Clinton ahead of Trump with probabilities of 82.3% to 16.8%.
• The average of opinion polls compiled by Realclearpolitics now shows an 8.6 point lead for Biden (51.1% versus 42.5% for Trump) after a 9.8 point lead on October 12th.
• According to FiveThirtyEight model simulations based on polls and demographic factors, however, Biden would win in 87 out of 100 simulations — up slightly from 86 nine days ago.
At the same stage in the 2016 campaign, Trump trailed Hillary Clinton by 5.1 points in the polls (43.2% to 48.3% according to Realclearpolitics, whereas 83% of the model simulations run by FiveThirtyEight had pointed to a victory for Clinton.
Current opinion polls show a wide dispersion depending on their source and methodology.
Democrats need more than a 3% margin!
Because of the Republicans’ advantage in the Electoral College, Biden would probably need to win the national vote with a margin of at least 3 points to carry enough swing states.
In key swing states, state polls now give Biden a lead of 3.5 points in Pennsylvania (down from 7.1 points nine days ago), a mere 1.6 points in Florida (it was 3.7 nine days ago), 6.3 points in Wisconsin (it was 5.5) and 3.1 in Arizona (it was 2.7) according to Realclearpolitics.
Biden would probably have to come ahead in either 1. Florida or 2. Pennsylvania — and one of the two smaller states for a majority in the Electoral College.
In the two major swing states, Florida with 29 and Pennsylvania with 20 votes in the Electoral College, the race has tightened in the last nine days.
Why 2020 is not another 2016
In 2016, Trump’s victory came as a surprise. While the 2016 race had narrowed significantly in the final two weeks of the 2016 campaign, the Realclearpolitics average of national polls had still indicated a 3.2% lead for Clinton on the eve of the vote with 46.8% versus 43.6% for Trump.
The final result with a 2.1% lead for Clinton had been within the usual margin of polling error.
The real surprise in 2016 was that state polls had underestimated support for Trump significantly in a few key swing states.
This allowed Trump to come ahead in the state-based Electoral College despite losing the national vote.
According to Realclearpolitics, this is unlikely to happen again in 2020. U.S. pollsters have refined their method, taking account of the correlation between education levels and support for Trump.
In addition, more quality pollsters are now surveying opinion more regularly in these swing states.
A new source of uncertainty
However, a new source of uncertainty casts a shadow over the upcoming election. In the age of COVID 19, more voters will cast their ballot by mail.
While there is no evidence to support Trump’s allegations that postal voting is prone to fraud, the share of postal ballots that are disqualified because of a missing formality or a missed deadline is higher than in standard votes.
This may potentially hurt the Democrats whose voters seem to be more wary of COVID 19 than many Republicans and who are thus more likely to vote by mail.
Will the U.S. Supreme Court pick the election winner?
The scenario of a “red mirage,” an apparent narrow win for Trump on election night in key swing states that vanishes once postal ballots are counted, refers to this.
More importantly, a narrow lead for Trump before postal ballots are fully counted could give rise to serious and extended legal complications.
If that happens, the 2020 election, as in 2000, may ultimately have to be settled by the Supreme Court if the Republicans try to stop the full count of postal ballots.
The good news: Lots of voting activity
According to the U.S. elections project, 37.9 million ballots have already been cast by now. That is far above the number at the same stage in the 2016 contest and 27.3% of the total 2016 turnout.
Judging by the states which provide detailed information, roughly 88% are postal ballots rather than early in-person votes.
In 2016, 1.2% of postal ballots were disqualified, mostly because of issues with the signature. In the Congressional election of 2018, the share rose to 1.4% with 425.000 of the roughly 30 million postal ballots rejected.
Senate: Democrats Ahead
Whereas the U.S. president sets foreign policy and has wide discretion in trade and regulatory policies, tax and spending decisions need to be passed by Congress.
U.S. policies after the election thus depend on the outcome of the parallel Congressional elections on 3 November.
While the Democrats will most likely keep their majority in the House (84.7% chance), the probability that they may wrest control of the Senate from the Republicans has slipped to 60.4% after 62.2% nine days ago according to electionbettingodds.com.
The final odds, with probability percentages
The US elections may thus yield three plausible outcomes:
Scenarios 1 + 2: Biden wins the presidency (60% probability) with or without a Democratic majority in the Senate (scenarios 1 and 2 with probabilities of 45% and 15%, respectively)
Scenario 3: Trump stays in the White House while Congress remains divided (40% probability).
Whereas Biden has the edge, the outcome remains open.
*chief economist at Berenberg Bank in London
**first published in: www.theglobalist.com