What Should You Make of These Insane Polls?
WHY YOU SHOULD CARE
Because you lived through 2016, and are now seeing some polls with Biden up 16 points over Trump.
By Nick Fouriezos
You are forgiven for feeling the whiplash. Last week was one of constant head turns, from Donald Trump’s incredibly low tax payments and an explosive debate in which the president refused to denounce white supremacists or promise to accept the results of a fair election — and that was all before he tested positive for COVID-19 and was helicoptered to Walter Reed hospital for treatment over the weekend.
You’re not crazy: Everything is just getting crazy around you. And it’s no different in the polling realm.
Democrat Joe Biden’s average lead over Trump in national polls has hardly shifted since the spring, mostly hovering between 6 to 8 points, remarkable in its consistency. Until the last few days, that is, when polls began spiking like they had just taken steroids and were suddenly feeling better than they had in 20 years. There was an NBC/Wall Street Journal poll that interviewed voters after the debate and had Biden up 14 percentage points; then there was the CNN poll released Tuesday that polled people after Trump contracted COVID — and gave Biden a 16-point margin.
That wild ride continued in key battleground state surveys where the race had been tight: Biden by 12 in Pennsylvania (Monmouth), by 8 in Arizona (New York Times/Siena). What gives?
Even in an unbelievable year, those numbers seem fantastical to most people, who took to social media to express their disbelief and urge fellow voters to not take their feet off the pedal. And after a week to process the results, the OZY/0ptimus Forecast — an exclusive predictive model that crunches polls, candidate traits, economic data and more to deliver a percentage chance of victory up and down the ballot — suggests people are right to be skeptical … but not too skeptical. Yes, the volatility of the election has finally delivered volatility among pollsters too. However, Biden is safer than one might expect — even if Election Day sees a polling blunder of epic (aka Hillary Clinton) proportions.
There was a polling error [of between 3 and 4 points] in most states in 2016. Even if that exact same polling error happened in 2020, Biden would still win — by a lot.
Scott Tranter, CEO of 0ptimus
“There is a saying in baseball: If you’re up five runs, not even a grand slam can tie it,” says Scott Tranter, the CEO and cofounder of 0ptimus, a data and technology firm. “There was a polling error [of between 3 and 4 points] in most states in 2016. Even if that exact same polling error happened in 2020, Biden would still win — by a lot.”
In the past week, a number of individual states have moved further away from Trump in polling. However, that hasn’t decreased Trump’s chances of winning significantly in the forecast, which is still around just 18 percent.
That is in part because Biden’s lead is ballooning in battleground states he was already favored to win, such as Wisconsin (Biden has a 76 percent chance of victory), Michigan (80%) and Minnesota (83%). Meanwhile, Trump has kept the polls just close enough to keep Florida and North Carolina competitive — with 35 percent and 44 percent of Trump winning them, respectively — which are key to him maintaining any sort of path to victory. “For Trump, at least those aren’t going away,” Tranter says, even if the others are moving out of reach.
The volatility in the polls would be more advantageous for Trump if they seemed to be going in both directions; that is, for every poll that shows a huge double-digit Biden lead, there was another one that showed a tight race. The week of the 2016 election for example, a Monmouth poll had Clinton up 6 points nationally … but then an LA Times poll had Trump up 3.
But the fact that some polls have Biden up by 16 points and others have him up only 5 suggests a variance that doesn’t necessarily mean Trump voters are lying in wait, ready to spring out for another Election Day surprise. “Volatility helps Trump as long as it occasionally jumps back Trump’s way,” Tranter says. “But volatility is only volatility before it becomes consistent — then it becomes a trend.” And given that those bumps happened after critical moments for the Trump campaign, from the tax revelation to the debate to the COVID diagnosis, there is a chance Biden is simply capitalizing, and these results are closer to the new polling reality. That would seem to be the case if the next few weeks see more polls with Biden having leads in the low- to mid-teens.
It’s important to consider more than the polls when assessing the election, which is why the OZY/0ptimus forecast considers a number of other factors — from historical trends and media coverage to economic indicators and fundraising numbers. For instance, the Trump campaign announced this week that he will be pulling millions in ads in Ohio and Iowa, two states that are currently toss-ups, one of many signs that his upcoming third-quarter FEC report is likely to show disappointing fundraising numbers for the incumbent.
However, the improving economy (on the heels of a somewhat positive September jobs report) should be a plus for him down the stretch — especially as many of the undecided voters OZY has spoken to in recent days indicate they will vote on their pocketbooks far more than the jaw-dropping news of the day. And the ultimate volatility will be the conduct of the election itself, as record numbers of absentee ballots could make for a chaotic count. For Trump, merely keeping the election close could be enough to throw it into doubt.
“Trump is doing the bare minimum to keep his head over water,” Tranter says. Or, to use another baseball analogy, given that the Democratic nominee can afford to lose a number of swing states and still win overall: Biden’s handing a solid lead to his bullpen, and Trump has precious few innings to mount a comeback.