For those feeling confident about a Clinton victory in the wake of recent polls showing Trump behind, consider the poll taken just before WWF wrestler Jesse “The Body” Ventura—a celebrity joke candidate with no political experience—“shocked the world” by winning the governor’s race in Minnesota in 1998. Ventura was polling third at 15 percent three weeks before the election but got 37 percent of the vote in a three-way race against Hubert Humphrey III, a well-liked Democrat whose family is as close to royalty as you can get in a liberal, deeply Democratic state notable as the only one to vote for Mondale in Reagan’s 49-state landslide in 1984. Minnesota Public Radio, which conducted the poll, reported Ventura as a footnote in the race that was seen primarily between Humphrey, the Democrat, and Coleman, the Republican:

“Attorney General Hubert Humphrey III enjoys a double digit lead over Norm Coleman . . . as the race enters its final three weeks. Statewide, 56 percent of voters have a favorable opinion of Humphrey, while 23 percent have an unfavorable opinion . . . In a match-up, Humphrey captures 44 percent of the vote, compared to 31 percent for Coleman, with 15 percent backing Reform Party candidate Jesse Ventura . . .”

What The Washington Post wrote about the race after the election sounds eerily familiar:

“Campaigning with little money for television ads but plenty of name recognition thanks to his wrestling, movie and talk radio careers, ‘The Body’ at first seemed not to take his own candidacy seriously. He openly spoke of resuming his radio show after Election Day. But as his cut-rate campaign gained in the polls, Ventura kept the experts on edge with his odd pronouncements – he publicly pondered the merits of legalizing drugs and prostitution – and redubbed himself ‘The Mind.'”

I suspect the polls are as wrong today as they were in Minnesota, for three reasons: (1) Polling firms are using outdated turnout models from 2008 and 2012 that fail to account for the tsunami of Trump supporters, many of whom are either voting for the first time or crossing party lines, and the relative indifference of Clinton supporters compared to Obama supporters in 2008 and 2012;  (2) Polls cannot account for the numbers of people who simply won’t admit to a pollster they’re voting for someone so outrageous as Trump; and (3) The WTHWN (“what the hell, why not”) factor. In the aftermath of Monica Lewinsky’s blue dress, George Bush’s trillion-dollar WMD fantasy, and premiums for Obama’s “Affordable” Care Act rising into outer orbit while coverage sinks beneath the sea, we simply no longer have the capacity to be shocked or disgusted by anything. That thinking is dangerous, as Trump so well proves, but understandable. 

The stealth factor behind Trump is well documented. He got more votes than any Republican in history in the primaries, but he consistently under-performed in polls.  According to Real Clear Politics, Trump polled at 48 percent on the eve of the Pennsylvania primary but got 58 percent of the vote. In Massachusetts, he did 5 points better in votes than he did in polls. In Maryland, his percentage of votes beat his polling average by a whopping 12 points.  If GOP primary voters in those states were too embarrassed to tell pollsters they were supporting Trump, imagine how bashful general election voters are. In Minnesota, Jesse Ventura’s candidacy was considered a sideshow, but his performance at the ballot box more than doubled his performance in polling three weeks before the election.  In other words, the polls very close to the election weren’t just wrong; they were wildly wrong.

You might be inclined to dismiss the Jesse Ventura shocker as an outdated anomaly involving a governor’s race that got far less scrutiny from top-level pollsters than a modern presidential campaign.  Perhaps—until you consider that Hillary Clinton led in every poll by double digits and 21.4 points in the Real Clear Politics average of all polls the night before she lost the Michigan primary to Bernie Sanders by 1.5 percent of the vote. That’s a 23-point polling error. Nervous yet? Perhaps you put your hope in uber-savvy pollsters like Nate Silver of FiveThirtyEight to know what’s coming. Well, Nate gave Hillary a 99-percent chance of winning the Michigan primary. Once again, the pollsters were wildly wrong. Why?

Bernie Sanders, Donald Trump, and Jesse Ventura all have one thing in common: each is more phenomenon than politician. Whatever you think of John McCain and Mitt Romney, no one could accuse either of them of being a phenomenon. Their losses were very easily and accurately predicted by pollsters. It’s harder to understand and therefore impossible to predict the behavior of a phenomenon. No one really knew why, much less predicted, millions of Americans would pay twice the minimum wage in 1975 for something called a “pet rock.” It was a totally new idea. It didn’t adhere to marketing norms that relied on things like utility, beauty, quality, and value to predict customer behavior. Pet rocks were unapologetically useless, ugly, and grossly over-priced, yet they sold like hotcakes. Likewise, we’ve never seen a major-party nominee like Donald Trump. He is unapologetically ignorant, crude, crass, inexperienced, unqualified, and over-hyped by every norm of American politics, and yet he may be our next president. Just ask Governor Ventura.

In Minnesota in 1998, Hubert Humphrey was well liked and well qualified. He led in pre-election polls but came in third. In America in 2016, Hillary Clinton is well-qualified but deeply disliked and distrusted. (When even Bob Woodward of Watergate fame says you are “corrupt,” you have a problem as a Democrat.) She leads the electoral college in every poll even as everyone, including those pundits publicly confident that she cannot possibly lose, privately worries that somehow she will lose—just as she lost to Bernie in Michigan when he couldn’t possibly win, and just as a respected, accomplished, trusted, well-educated, experienced and much-loved son of a Democratic icon in an overwhelmingly Democratic state lost to an inexperienced, self-promoting buffoon in wrestling tights.

The truth—and it’s high time we admitted it—is that no one really knows what’s likely to happen on November 8, which is why everyone is on edge. Predicting where Trump will do well and by how much is like trying to predict the track of a hurricane. But in my unstudied opinion (and one I dearly hope will be disproved), anywhere that Clinton is not ahead in the polls by at least five points, it is safe to assume she is tied or slightly behind in the actual vote.

I mailed my absentee ballot for Clinton from London a month ago. The three pounds and seventy pence it cost me for postage was money gladly spent, but I have a sneaking suspicion it will be to no avail.  Look for the polls to tighten in the days ahead. Trump not only can win this thing, he is winning, and all this complacent talk about an inevitable Hillary Clinton presidency is not helping.