Hello. It looks like you’re using an ad blocker that may prevent our website from working properly. To receive the best Tortoise experience possible, please make sure any blockers are switched off and refresh the page.

If you have any questions or need help, let us know at memberhelp@tortoisemedia.com

TOPSHOT – Sri Vasamsetti, 22, of Seattle and a supporter of Democratic presidential candidate Hillary Clinton, watches televised coverage of the US presidential election at the Comet Tavern in the Capitol Hill neighborhood of Seattle, Washington on November 8, 2016. / AFP / Jason Redmond (Photo credit should read JASON REDMOND/AFP via Getty Images)
Sensemaker, 29 October 2020

Sensemaker, 29 October 2020

TOPSHOT – Sri Vasamsetti, 22, of Seattle and a supporter of Democratic presidential candidate Hillary Clinton, watches televised coverage of the US presidential election at the Comet Tavern in the Capitol Hill neighborhood of Seattle, Washington on November 8, 2016. / AFP / Jason Redmond (Photo credit should read JASON REDMOND/AFP via Getty Images)

What just happened

One reason people like polls is that they provide comfort – they prepare us for what’s coming. Except when they don’t. Given how often they’ve been wrong recently, how confident should we be that Joe Biden’s polling lead will translate into victory?

There are five days to the US election, and this is what we know: more than 75 million people have already voted. That’s about 56 per cent of the total number of voters in 2016, driven by an unprecedented number of postal votes. National polling averages put Biden roughly ten points ahead of Donald Trump, at 52 percent to 42. FiveThirtyEight puts the chance of a Biden presidency at 88 percent; the Economist96.

At state level, these national polls translate into a large number of states being up for grabs for Biden, including Texas – which hasn’t voted for a Democratic presidential candidate since Jimmy Carter in 1976.

So far, so familiar

But what’s to stop the same polling failures we saw in 2016 happening all over again? Well, a few things are different this time:

  • The Democratic lead in the prediction models is simply bigger this year than in 2016. For example: at this stage in the 2016 election, Hillary Clinton had a 66 per cent chance of winning, according to FiveThirtyEight’s model.
  • As a result, even with the same polling error as in 2016, key battleground states would still go to Biden, handing him the 270 electoral college votes he needs.
  • Also, there is little sign of the last-minute shift towards Trump by swing voters seen in 2016.

But that’s not the only difference

  • Pollsters have also learned from their mistakes, among them a failure in 2016 to take sufficient account of voters’ education levels, which turned out to be crucial for accurate predictions. College-educated Americans, who tend to be Democrat-leaning, were more likely to take part in polls. Several pollsters, including Florida’s UNF and Pennsylvania’s Muhlenberg College Institute of Public Opinion, have re-weighted this variable in their models. Re-running UNF’s final pre-election poll in 2016 with this new weighting flipped the result from a 5 per cent lead for Clinton to a 2 per cent lead for Trump. Trump won Florida that year by 1.2 per cent.
  • The other group that jumbled results in 2016 was undecided voters, and there are far fewer of those this time around. It turned out that a lot of undecided voters opted for Trump at the last minute, something that pollsters assumed wouldn’t happen.
  • Some have dared to hint that, scarred from their previous wounds, pollsters may have overcorrected on these fronts.

Reasons to be wary

  • Pollsters use different methodologies and may have an agenda. This week the Trafalgar Group published a poll purporting to show that Trump was going to win 30 per cent of declared Democrats’ votes in Michigan. When Nate Silver’s FiveThirtyEight website pointed out that this was absurd, the Trafalgar Group’s supporting data mysteriously disappeared. And yet all these pollsters’s findings get absorbed one way or another into polling averages produced by forecasters – like Silver.
  • On the flip side, pollsters can be scared to publish a weird-looking result. This can lead them to actively fiddle with their numbers or refuse to publish a poll if it looks out of whack compared with others. This means polls can counterintuitively become less reliable closer to election day as pollsters fret that they will be embarrassed by big polling errors. In a recent example, a leading Australian pollster ripped up a poll that seemed to show the Australian Labor Party losing the election – only for it later to be proved accurate.

Known unknowns

  • Early voters have heavily leaned Democrat so far, but we have known for months that they are more worried about Covid than Republicans and more likely to vote early. The big question is how much of a comeback Republicans will mount on the day.

Trump’s path to victory

Trump can win. FiveThirtyEight says he has a 12 percent chance, which is one in eight. That’s about the same as the chance of a train being more than three minutes late in the UK – unlikely, but you wouldn’t bet on it. You certainly wouldn’t buy a train ticket that relied on a three-minute connection.

Pennsylvania, where Biden currently has a five-point lead in the polls, looks like this year’s tipping-point state. Five percentage points is above a normal margin of error, but only just, and if Trump wins Pennsylvania his chances of a national win would jump to 69 percent according to FiveThirtyEight.

In this scenario, assuming a substantial polling error favouring Trump, he would start feeling much more comfortable in key sun belt states that are currently tipping towards Biden – including Florida and Arizona. Taking these states could turn the tables for Trump and win him 270 electoral college votes even if Biden wins back the rust belt states that turned against Clinton four years ago.

Long stories short

  • France will enter its second national lockdown tomorrow, which will last until at least the end of November.
  • Joe Biden confirmed that, if he wins next week’s election, he will look to pass a new Equality Act within his first 100 days in office, enshrining rights for LGBTQ+ people.
  • When he isn’t campaigning against the British government, Marcus Rashford also plays football. The Manchester United forward scored his first ever club hat-trick last night.

In the app today… As part of our ongoing File on the spiritual battlegrounds ahead of next week’s US presidential election, we have published a detailed analysis of Amy Coney Barrett’s recent testimony to the Senate Judiciary Committee. What does the new Supreme Court justice really think? And don’t miss our ThinkIn at 6.30pm: Matt d’Ancona in conversation with the presenter and writer James O’Brien. They – and our members – will be discussing the difficult art of changing your mind.

Please share this Sensemaker with your friends and colleagues.

Wealth investment, fairness, prosperity

Expensive fractions
There have already been some strange ol’ business deals this year: think Oracle’s very political partnership with TikTok in the US. And now there’s another example to add to the list: the French company LVMH’s acquisition of the upmarket jewellers Tiffany, the biggest ever such deal in the luxury goods market. As recently as a few weeks ago, the whole idea seemed to have been scrapped, as the two sides litigated against each other. But now, overnight, it appears to have gone through – with a $15.8 billion price tag. The clincher? LVMH has succeeded in making Tiffany drop its price per share, against a backdrop of worsening business projections during the pandemic, so that the overall cost is 3 per cent lower than it would otherwise have been. Hang on: three per cent?! That adds up to millions of dollars, but it’s still not a massive saving after all the months of argument and recrimination.

New things technology, science, engineering

Trimmed down to size
Jack Dorsey, the founder of Twitter, has grown a big, straggly lockdown beard. It made its first major public appearance yesterday when he himself appeared, via video-link, before a US Senate committee – and it caused quite a stir. Predictably, it now has its own Twitter account. But there’s a serious point behind the hirsute hilarity: the past couple of years have seen the tech titans dragged more and more into the public eye. The Senate hearing was about Section 230, a controversial provision in American law that effectively absolves social media companies from responsibility for the “third-party” content that’s published on their sites, and it also featured Facebook’s Mark Zuckerberg and Alphabet’s Sundar Pichai. Dorsey came under the heaviest questioning, though, particularly from the committee’s Republican members – although this kind of attention isn’t going to disappear if the Democrats take over next week.

Our planet environment, natural resources, geopolitics

It would loom over the Empire State Building, dwarf the Eiffel Tower, and embarrass The Shard. No, not the Burj Khalifa, but a 500m-high coral reef found in Australia’s Great Barrier Reef. Its emergence only now suggests just how much of the ocean remains unexplored (more than 80 per cent, according to the US National Ocean Service) and just how large the Great Barrier Reef is (covering an area 40 per cent bigger than the UK). A team of scientists discovered the reef while conducting 3D-mapping of the seafloor in the area. They are on a roll: back in April, the same team also discovered the longest recorded sea creature, a 46-metre siphonophore, twice as long as a humpback whale. But against the wonder of new discovery, a reality check: we’ve lost half of our coral reefs in the past 20 years.

The 100-year life health, education, living, public policy

The latest prognosis
Covid-19 infection rates across England are growing rapidly, according to the latest results of Imperial College London’s REACT-1 study. Their data suggest that 96,000 people are infected every day and that number is doubling every nine days. Professor Steven Riley, one of the authors of the report, told the BBC that current measures to contain the spread of the virus are “not sufficient” and that there “has to be a change before Christmas”. It’s sobering news but not unexpected. What we need now is quick and decisive action from government, otherwise we’re staring down the barrel of a winter crisis with rising hospitalisations and deaths.

Belonging identity, society, beliefs, countries

Enough is enough
“I won’t survive it, I know. Not this time. Enough is enough, I can’t do more. I am so desperate.” Those are the words of Nazanin Zaghari-Ratcliffe, a British-Iranian woman imprisoned in Iran since 2016, who faces a return to prison after months under house arrest in Tehran. Iranian authorities will try Zaghari-Ratcliffe under new charges on Monday, after which she will be incarcerated once again. This case has been brought just days after UK courts delayed a hearing about the country’s £400 million debt to Iran over a cancelled arms deal, although both countries deny the link. Last year, Tortoise spoke to Zaghari-Ratcliffe’s husband Richard about the family’s ordeal. And in September we reported on the brutality of Evin prison, where she has been held for most of the past four years. At the centre of this diplomatic issue is a desperate family. Zaghari-Ratcliffe’s return to them is long overdue.

Thanks for reading, and do share this around.

Kim Darrah

Charlotte Soin

Additional reporting by Xavier Greenwood, Ella Hill and Peter Hoskin.

Photographs Getty Images and Schmidt Ocean Institute


In yesterday’s Sensemaker email, one sentence incorrectly implied that Apple was paying money to Google to have Google as the default search engine on iPhones – when, in fact, the money has gone the other direction. This has been corrected in the online version of yesterday’s Sensemaker.

Some readers couldn’t click through to yesterday’s Opinion piece by James D. Zirin on the appointment of Amy Coney Barrett to the US Supreme Court. Here is a working link.