Driving across the country last week, it seemed hard to believe an American presidential election is happening a week from Tuesday. Few campaign signs sprout from urban lawns; partisan billboards along the highways are scarce. Away from the coasts, the talk on the radio is largely of football and Jesus, not politics. It takes a moment, hearing a spot in North Carolina for a US Senate candidate, to realize the voice belongs to President Obama, interrupting some country music.

Oh, there’s plenty of chatter about it in the raging echo chambers of talk radio and TV cable news, and in the cocksure journalists’ fun house known as Twitter, where in-the-tank reporters and dispossessed campaign consultants, smarting over their collective defeat in the primaries, smugly assure each other that Donald Trump will lose in a landslide.

But what if the widely swinging polls, turnout models and forecasting mechanisms are all wrong? What if the unique historical circumstances of this election — pitting the female half of a likely criminal family dynasty against a thin-skinned bull-in-a-china-shop businessman — have invalidated conventional wisdom? What if the ranks of shy voters storm the polls and, in the words of Michael Moore, deliver the biggest rebuke in history to the establishments of both parties?

What if, far from having a lock on 1600 Pennsylvania Ave. come January, Hillary Clinton’s margin-of-error lead — currently between 4 and 5 points in the RealClearPolitics average of multiple national polls — turns out to be a Potemkin village, dependent on high turnout among blacks and other minorities and on getting late deciders to turn her way?

What if, in fact, the opposite happens — that Trump’s appeal to the disaffected white working class (many of them Democrats) in coal-mining and Rust Belt states outweighs the Democrats’ traditional advantages in the big cities, flipping a state like Pennsylvania from blue to red?

Welcome to the hidden election, where those who say they know what’s going to happen don’t, and those who do know will make their voices heard on Nov. 8.

Nationally, Clinton holds 3.8 points over Trump in a four-way race that also includes Libertarian Gary Johnson and the Green Party’s Jill Stein. But polls may not be everything this year.

Indeed, Hillary has suffered a major polling meltdown over the past week or so, hurtling from a 12-point lead to 4 points in the Washington Post-ABC News tracking poll.

    

The WikiLeaks revelations about her campaign’s dirty tricks, the pay-to-play nature of the Clinton Foundation, the astonishing personal enrichment of the Clintons via politics and the electrifying news Friday afternoon that the FBI is reopening its investigation into her use of a private e-mail server are finally taking a toll.

Spurning the poll-based forecasts in favor of historical analysis, professor Helmut Norpoth at SUNY Stony Brook — who’s correctly predicted the last five presidential elections — gives the nod to Trump, 52.5-47.5 percent. Meanwhile, an artificial-intelligence system developed in India that takes into account data from Google, YouTube and social media says Trump’s “engagement data” points to a GOP victory.

So, if the conventional wisdom is wrong, what’s Trump’s plausible path to 270 electoral votes? In 2012, Mitt Romney won 206 electoral votes to Obama’s 332. But recall that the Electoral College is a zero-sum game; every vote that switches is both a plus and minus, so that’s not quite as big a margin as it might seem.

Current thinking has it that there are 11 battleground states that could go either way: Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania, Virginia and Wisconsin. In these states, with a total of 146 votes, the election will be won or lost.

Throw out the first tier of Colorado (nine electoral votes), Michigan (16) and Wisconsin (10), which are likely to stay blue; in the RealClearPolitics poll averages, Hillary Clinton leads by 6.2 to 8.8 points in this group.

A second tier would include Iowa (six), Nevada (six), and New Hampshire (four); of this group, Trump currently leads only in Iowa, by 1.4 points. But a win in any one of these could well provide a crucial margin of victory for him after the main battlegrounds of Florida (29), North Carolina (15), Ohio (18), Pennsylvania (20) and Virginia (13).

Because here’s the good news for Trump: Despite the structural advantages in the Electoral College the Democrats currently enjoy — they start with New York (29), Illinois (20) and California (55) already in their pockets — the truth is that Trump need only retain the states Mitt Romney won in 2012 (including, critically, North Carolina) and then flip these three battleground states: Ohio, Pennsylvania and Florida. That would give him a 273-265 victory.

Right now, the RCP numbers show Clinton up 0.7 points in Florida, Trump up 1.1 in Ohio, and Clinton up 5 points in Pennsylvania. Still, RCP has just put Pennsylvania into the “toss-up” category, including Arizona, Georgia, North Carolina and even Texas. And it’s likely that those four red states will remain true to form, barring a complete Trump collapse in the last week of the campaign.

Let’s take a look at each:

Florida

In Florida, Hillary will have the usual Democrat advantage in and around Miami and the university towns of Gainesville and Tallahassee. And the GOP’s hold on the Cuban émigré vote has softened over time, especially in the aftermath of Obama’s normalizing of relations with the Communist country. But Trump benefits from a certain favorite-son status (he owns Mar-a-Lago, in Palm Beach), has extensive business interests in the state and will easily win the rural and western parts of the state, while siphoning off some Jewish votes due to his adamantly pro-Israel stance. Clinging to a mere 0.7 percent lead in the RCP aggregates, Hillary’s head cannot rest easy in the Sunshine State.

Down-ballot leading indicator: Incumbent Sen. Marco Rubio is up 5.6 points over challenger Patrick Murphy.

Prize: 29 electoral votes.

Ohio

In Ohio, Trump has successfully wooed the working class, so much so that the Clinton campaign has pretty much given up on the bellwether state. Only one state poll shows Clinton with a small lead, while others give it to Trump by up to 4 points. Ohio is the very model of a Trump state, with a large industrial working class that’s seen jobs exported and its livelihood threatened; sure, Cleveland will roll over for Hillary, but look for Trump to be strong elsewhere, especially in the coal towns along the Pennsylvania border.

Down-ballot leading indicator: Incumbent GOP Sen. Rob Portman is demolishing former governor Ted Strickland, leading by almost 16 points in the RCP averages.

Prize: 18 electoral votes.

Pennsylvania

    

Things are tougher in Pennsylvania, a state that performs a quadrennial fan dance to tease the GOP, but then reverts to type as after-hours votes from Philly and Pittsburgh come flooding in. The RCP averages show Clinton with a healthy 5.2-point lead, but savvy observers know you have to also figure in a small but significant vigorish for the Democrats as last-minute poll irregularities are discovered, Democrat lawyers get emergency injunctions to keep polling places open and boost minority turnout, and ballots fall off trucks or are discovered in locked rooms. A Trump wave, especially among disaffected Dems and outsourced steel workers, could flip the state, but it will still be hard.

Down-ballot leading indicator: Incumbent Sen. Pat Toomey clinging to a 1.3-point lead over challenger Katie McGinty in a state that often retires new GOP hires after one term.

Prize: 20 electoral votes.

If Trump loses Pennsylvania, his next-best chance to close the deal comes from Virginia and Iowa/Nevada, where the combined 19-25 electoral votes would just squeak him over the line. But thanks to the metastasizing numbers of federal employees in the northern Virginia DC suburbs, the Old Dominion is no longer a sure thing for the Republicans; Hillary is currently up 8 points. Neither is Nevada, Harry Reid’s service-employees fiefdom, where Hillary leads by 2 points.

Meanwhile, consider this: If Trump loses Pennsylvania and Virginia, but (in addition to Ohio and Florida) wins New Hampshire, Iowa and Nevada, the race ends in a 269-269 tie. Then it might come down to a single congressional district in Maine or Nebraska (neither a “winner take all” state) or even wind up in the House of Representatives.

As the campaigns draw to a close, however, it’s crucial to remember that polls are only one aspect of the race. Still little understood is Trump’s lead in the unconventional metrics of social media. Consultants and reporters like to crow that rally sizes (huge for Trump, miniscule for Clinton unless she has Michelle Obama by her side) are not predictive of electoral success.

There’s little question that the Trump-Pence ticket has generated far more visible enthusiasm among its supporters than the dour Washington death march of Clinton and Tim Kaine.

Less remarked is Trump’s overwhelming superiority on social media, including Twitter, Instagram and Facebook, where Trump’s page has double the “likes” of Clinton’s (10 million to 5 million). Further, Trump is said to average 30,000 viewers on live-streamed YouTube events, compared to Clinton’s mere 500.

Clinton has big money on her side, but Trump has big motivation. Both the “intangibles” and AI models showing a Trump victory take these factors into account.

The truth is, this is an election not just between Clinton and Trump but a whole raft of political antagonists in Barack Obama’s “fundamentally transformed” America: urban vs. rural; old vs. young; makers vs. takers; taxpayers vs. recipients; white collar vs blue collar; Harvard vs. the heartland; manipulative consultants and biased reporters vs. honest Americans who, however naively, believe that their vote really does matter.

Many have felt apathetic or disenfranchised for decades.

The question is: How many of them are there and are there enough of them to hold the GOP line and deliver the three crucial states to Trump? We’ll soon find out.

About the author

Michael Walsh is an author, screenwriter and contributing editor at PJ Media. His most recent book is “The Devil’s Pleasure Palace.”