Less than week out from the 2016 Presidential election Hilary Clinton was 2.8% ahead of Donald Trump in the 538 average of polls. Six days out in 2020 Joe Biden is ahead in the 538 average by 8.8%.
The RealClearPolitics average gave Clinton a 3.2% lead. Joe Biden, at the same point in 2020 has a lead of 7.1%.
More than 70 million votes have already been cast – a record number – and it appears that turnout might be the highest for a century.
RCP and 538 rate all polls for accuracy but include the findings in the average. Thus Rasmussen polls, which consistently show Trump in the lead nationally, get in and have a marginal impact on the averages.
The Washington Post, a week out from the election, said their average of high quality polls gave Biden a 9% lead with Biden up 7% in Pennsylvania; 8% in Wisconsin, 9% Michigan, 5% Arizona, 4% North Carolina and 1% in Florida. Another Post poll had Biden ahead in Wisconsin by 17% but that is probably an outlier.
538’s averages for battleground states have Biden up in Arizona by 3.5%, Florida by 2.3%, North Carolina by 2.1%, Pennsylvania by 5.1%, Michigan 8%, Georgia, 1.4% and Iowa 1.5%. Texas is line ball but unlikely although by 2024 it may well be a Blue State. The 538 probability estimate gives Biden an 88% chance of winning the Electoral College.
RCP averages for battleground states has Biden up 6.4% in Wisconsin, 0.7% in North Carolina, 3.8% in Pennsylvania, 8.6% in Michigan, 2.2% in Arizona with North Carolina, Florida and Georgia ties.
The Economist has a different system combing State and national polls and economic indicators. It says there is a 99% probability that Biden will win the popular vote and 97% probability that he will win the Electoral College estimating a 357 to 181 win for Biden. It’s average of polls gives a 9% lead to Biden.
The Guardian average of polls gives Biden a 9% lead with Biden up in Florida 1%, Pennsylvania 5.3%, Michigan 7.3%, Arizona 3.6%, North Carolina 2.2%, Wisconsin 6.7% and Iowa 0.3%.
The Financial Times average of polls gives Biden an 8.6% lead. It focusses on probabilities in individual States to come up with a figure of Biden having 207 solid Electoral College votes and 72 States leaning towards him. Trump has 83 solid states and 42 leaning. It puts 134 in the toss up category. If Biden wins all the States leaning towards him he is elected with 279 Electoral College votes. If Trump wins all the toss-ups plus all those leaning to him he still falls short.
And an insight into one small section of the population, the scientific community, comes from a Nature poll on issues and presidential preferences. Readers rated climate change and the pandemic as their chief worries in the election. Of 892 respondents in total, 86% supported the Democratic candidate, 8% favour Trump and 6% responded ‘Other’ with some writing in Bernie Sanders and libertarian candidate Jo Jorgensen.
Needless to say there is also lots of anecdotal evidence much of it useless. But there is an indicator of campaign priorities in which areas the candidates are travelling to in the last week or so. In a Washington Post piece (28 October) Karen Tumulty, she said: “If candidates and their top surrogates show up in unexpected places, it can be a sign for good or ill. Right now, indications suggest President Trump’s team sees itself in a dire situation, while former vice president Joe Biden’s is sensing not only a victory but also the possibility of a rout.”
Of course, as everybody is saying, Donald Trump might still win thanks to voter suppression; unwillingness of Trump supporters to respond to polls or to lie; electoral fraud; and, the Supreme Court gifting victory to him.
But, if you are suffering post-2016 existential angst about the outcome The Washington Post’s Phillip Bump (24 October) had a useful guide to how under or over estimates of Trump’s opinion polling effect likely Electoral College outcomes which might alleviate it slightly. At that stage it would take an underestimate of 5.5% for Trump to win. You can adjust his gauge to explore the various options.
Also keep in mind the old story about Einstein and Morgenstern helping Kurt Godel with his application for US citizenship. Godel told them that in preparation he had been studying the US Constitution and had seen how the US could become a dictatorship. They urged him not to bring the subject up in his interview.
Scholars think Godel was wrong but that he may have been thinking about Article V. Trump has probably never have heard of Godel, nor Article V, but he obviously shares his belief having claimed that as President he can do whatever he likes.
And before you scoff at the prospect, polls indicate that between 14% and 18% of Americans (left and right combined) believe some violence may be used to achieve a political goal. The Economist (17 October) reported that in September 16.7 million firearms were bought – more than any full year on record. Ammunition sales were 66% higher than a last year.
Only in America as the saying goes. So what’s the most likely outcome? If you wanted to have a bet the odds returning the most would probably be those on a Trump win (with or without Supreme Court intervention) with the second best odds a Biden landslide.