few hours ago I wrote a piece explaining in some detail why, if you take a 360-degree view of all of the polling data we’ve gathered over the last four years, the current view—that Donald Trump is losing by roughly 9 points—makes sense.
Since the time of that writing there have been two developments that suggest we can take this conclusion further:
Donald Trump is going to lose the presidency, probably by a historic margin.
The only thing that could save Trump in this environment would be a COVID-driven black swan in turnout. And for Trump that would mean low turnout. Because the lower the turnout, the less representative the polls might be. The higher turnout, the greater the polls’ predictive power.
Which brings us to the second item: FiveThirtyEight is predicting a total turnout between 144 million and 165 million votes—with their pin set at 154 million votes.
It was difficult enough to see Trump winning with turnout under 140 million votes. But if we wind up north of 144 million it becomes basically impossible.
One reason turnout matters is that the higher it goes, the less important the samples are—and thus the more reliable the polls are as indicators of reality.
Put those two numbers together—Trump -13 on reelect and turnout in the vicinity of 154 million—and you have a campaign that is effectively over. The only open question is the magnitude of Trump’s coming loss.