‘WHO you gonna believe, me or your own eyes?’ asked Chico Marx in Duck Soup.
And with polling day in the US presidential election upon us, the gap between what the polls are telling you to believe and everything else keeps growing. Donald Trump is going to lose. On FiveThirtyEight, Trump is 8.5 points behind in the national average of polls and RealClearPolitics has him 7.2 points behind.
However, all the other evidence indicates strongly that Trump is going to win. Most people feel better off than four years ago, Trump’s rallies attract massive attendance in swing states such as Pennsylvania and newspapers such as the Pittsburgh Post Gazette have endorsed a Republican candidate for president for the first time since 1972. The bookies still have Trump at 15/8 to win but 58 per cent of all individual bets have been on Trump winning, with 62 per cent of bets in October on Trump to win; the closer one gets to polling day, the greater the number of people putting their money on Trump despite what the polls are telling them.
A closer look at the polls shows inconsistencies that demand an explanation. Opinium gave Biden a 14-point national lead but the Democracy Institute poll puts Trump 1 per cent ahead and likely to win in the electoral college 326-212 as his support is spread more ‘efficiently’ than Biden’s, which is heavily concentrated in coastal states. The Democracy Institute’s track record is excellent, correctly calling the last US election and the EU referendum. They accept that there are a significant number of Trump supporters who will not publicly state their allegiance, whether through fear or bloody-mindedness and ask questions to quantify this reluctance to reveal their true voting intentions. Their current poll found 79 per cent of Trump supporters would not admit their voting intention to friends or family, whereas only 21 per cent of Biden supporters felt this way. The Opinium poll implicitly acknowledges the existence of shy/silent Trump voters in its detail, with Biden leading by only 4 points when voters are asked who they think will win.
Some pollsters have come to the conclusion that they don’t believe their own polls. The head of the Trafalgar Group’s prediction is that his ‘best guess would be an Electoral College victory in the high to 270s, low 280s’ for Trump. USC Dornsife’s latest poll shows Biden ahead by 11 points but the result is much closer when they ask people who their friends are voting for, so much so that USC Dornsife predict Biden will win the popular vote but lose the electoral college.
Covid has increased the number of mail-in (postal) and early in person ballots cast, making comparisons with 2016 near-impossible. Stories abound of mail in ballots not being sent, millions not yet returned, or declared invalid. This is unlikely to be helping Democrats, who made a big early push for people to vote by post whilst Republicans have urged their supporters to vote in person on polling day if at all possible. Both sides have lawyered up for the court battles that look increasingly likely in any state with a close result. Pennsylvania will accept mail in ballots three days after polling day, a decision that the Supreme Court chose – splitting 4-4 – not to suspend, but thereby leaving open the possibility of further legal challenge. With Amy Coney Barrett now installed on the court, expect Democrat howls of fury should she be the deciding vote on any electoral challenge.
Trump has the advantage in momentum, enthusiasm and perhaps now in the courts. The polls are against him (mostly) but the pollsters are not sure. So there you have it: Trump to win, maybe probably.