HARRISBURG, Pa. —  Wes Bair, a 74-year-old Black man from Pennsylvania’s capital city, has personal experience with young Black voters wavering on whether to support Kamala Harris for president.

“Even my son was saying, ‘What’s she done?’” Bair said as he waited in the shade on an unseasonably warm day this week to get into a rally here for the vice president.

But, he added, “I’m on team positive.” As the remaining fence-sitters focus on the major candidates and their positions, Harris will gain the backing she needs, he predicted.

The night before, he said, his son and daughter-in-law called.

“We thought they were going to tell us that she’s pregnant,” Bair said. But it was a different sort of welcome news: “Mom and Dad, we’ve decided we’re voting for Harris.”

Multiply that by a few million, and you have the tension of the final week of what polls show is one of the closest presidential races in American history:

Across the U.S., especially in the seven states where the outcome appears up for grabs, two opposing armies face off, battling over a relative handful of people still uncertain about whom to support, or more commonly, whether to vote at all.

In an agonizingly close race, the result may pivot on the individual decisions of that tiny slice of the electorate — mostly voters who have tuned out the cacophony of the campaign until recently and are only now coming to grips with their decision.

How can the race still be so close?

I can’t tell you who will win. Polls are simply too close.

As of Friday, the average of nationwide polls maintained by 538 showed Harris ahead by 1.2 points. In five of the seven swing states, 1 point or less separated the two candidates.

This is driving many Democrats crazy. How, they want to know, can the election remain so tight?

But it’s Republicans who should be asking that question.

Incumbent parties in wealthy democracies have been thrashed around the world over the last two years.

Angered by the worldwide bout of inflation that followed the COVID-19 public health emergency, and a huge increase in the number of migrants fleeing war, dictatorships and the impacts of climate change, voters in Europe, Asia and the Americas have turned against their political establishments. That’s been true for parties of the center-left, including Justin Trudeau’s Liberals in Canada, and the center-right, such as Britain’s Conservatives.

The same forces shape politics in the U.S.

Michigan State University political scientist Matt Grossman recently noted a consistent phenomenon in models designed to forecast the election outcome: Those that take polls into account have tended to show a slight edge for Harris. But models that ignore the specific candidates and look only at what political scientists call “fundamentals” — such as the rate of economic growth and voters’ approval of the incumbent president — have generally pointed to a Republican win.

The takeaway is simple: A normal Republican would be planning a victory party right now.

Instead, former President Trump is locked in a toss-up race.

That partly reflects how well Harris has put together a campaign in the little more than three months since President Biden left the race. Mostly, however, it’s a measure of how Trump’s unpopularity holds down Republican fortunes.

Some crucial questions to consider

The macro-question in these final days before the election is whether Trump’s unpopularity will sink the GOP despite the worldwide tide that has been lifting up opposition parties.

That, in turn, boils down to a series of other, interrelated questions that will largely determine who wins. Here are four:

• How big will Republican pickups be among voters of color, and will they be offset by Democratic gains among college-educated white voters?
• How large will turnout be, and how much of it will be made up of people who seldom show up for elections?
• Will Black and Latino voters who say they support Trump, especially young men, actually show up at the polls?
• How much has the Republican advantage in the electoral college shrunk?

In the closing weeks of the campaign, Harris has tried to pull off a tricky balancing act: aiming to maximize her appeal to college-educated suburban voters, including Republicans and Republican-leaning independents, while seeking to minimize losses among young voters of color. Those groups have different, sometimes opposing, priorities.

Trump, by contrast, has pursued a more one-dimensional approach: doubling down on appeals to men.

Shifting racial and gender voting patterns

Which strategy will be more effective is hard to know in advance, in part because high-quality polls differ on what the 2024 electorate will look like, even as they agree that the race remains very tight.

Some surveys, notably the New York Times/Siena College polls, point toward a significant realignment of long-standing racial voting patterns, with Harris gaining ground among college-educated white voters, especially women, even as Trump builds support among Black and Latino men, especially those without college degrees.

Other surveys show a smaller gender gap and fewer racial differences

If the NYT/Siena polls are correct, the country is becoming less racially polarized, but more divided by gender and education.

That would be a blow to a cherished belief among many progressives — that voters of color will consistently put racial and ethnic identities ahead of other interests and side with candidates of the left. Instead, Democrats may become even more reflective of the political priorities of college-educated white voters, who form their largest base of support.

It’s not clear which party might gain a net benefit from such a shift.

White voters’ share of the electorate has slowly decreased, but in 2020 they still made up 72% of voters. If Harris gains even a little ground among them, that could offset declines among Black and Latino voters, who made up 11% and 10%, respectively, according to data from the Pew Research Center’s detailed study of who voted that year.

How much change will this election bring to the country’s racial and gender divides? The answer will not only go a long way toward deciding the next president — it could also set the pattern for years to come.

Which voters, and how many, will turn out?

The 2020 election saw 158.7 million votes cast — the highest turnout in American history, both in raw numbers and as a percentage of the adult population. This year seems likely to at least come close.

Gallup finds 70% of registered voters saying they’re “more enthusiastic than usual” this year about voting — about the same level as four years ago, and noticeably higher than in 2016 or 2012.

Democrats, who were less enthusiastic than Republicans earlier in the year, now have an edge on that measure, with 77% of those surveyed saying they’re more enthusiastic, compared with 67% of Republicans, Gallup found.

Despite that enthusiasm edge, high turnout doesn’t necessarily mean a Democratic advantage the way it once might have.

College-educated voters tend to consistently show up. As Democrats have gained ground with them, the party has racked up a string of victories in low-turnout special elections and off-year contests.

By contrast, Trump’s base includes a lot of people who don’t routinely vote. And those he’s been courting most heavily in the campaign’s closing stretch — young men — have some of the lowest turnout rates of any group.

A new Monmouth University poll of Pennsylvania, the largest and most closely fought of the swing states, showed Harris leading 51% to 46% among voters who cast ballots in all of the last five federal elections. Trump, by contrast, led 47% to 42% among those who voted in none or just a few of those elections.

In both 2016 and 2020, Trump showed an ability to inspire high turnout among infrequent voters, especially the rural white conservatives who are his core supporters. This election, he’s hoping to add on younger and infrequent male voters.

As Democratic pollster and strategist Anna Greenberg recently said on X, “the Harris coalition rests on the most reliable voters (older, college educated). Trump needs every single low propensity less educated young person to come out and vote for him.”

Literal trash talk, Bad Bunny and voters afloat

One particularly important aspect of the turnout question involves the Black and Latino voters who tell pollsters they back Trump.

All year, analysts have raised questions about how many of them will actually show up on election day, especially in the face of the incendiary racial politics that Trump employs.

Voters who flirt with a new party often come home to their traditional moorings in a campaign’s final weeks.

Sometimes, events can accelerate that.

On Sunday, a comedian at Trump’s rally at Madison Square Garden referred to Puerto Rico as “literally a floating island of garbage in the middle of the ocean.”

Democrats quickly denounced the insult. Republicans backed away from it. Celebrities from the U.S. territory, including the popular recording artist known as Bad Bunny, publicly stated their support for Harris.

In Pennsylvania, Puerto Ricans make up about 8% of the population, according to census data. That’s the fourth-largest share of the territory’s residents among U.S. states. And as American citizens, they’re eligible to vote. Most typically vote for Democrats, but many don’t vote at all.

Jose Rivara, a 30-year-old resident of Hershey, about 15 miles east of Harrisburg, said that before the “island of garbage” controversy, the fellow Puerto Ricans he knows had been thinking more favorably of Trump.

“Illegal immigration is the No. 1 thing they’re upset about,” he said.

Many of them have heard that new immigrants are “getting eligibility for government support” for housing and work permits, and believe the newcomers are obtaining help that U.S. citizens don’t receive, he said.

Rivara said he was living in Puerto Rico when Hurricane Maria hit in 2017.

“Trump delayed government funds” for the islands, he said, and people died as a result.

“I thought that was really heartless,” he said, adding that it settled any question in his mind about which candidate to back. And now the “garbage” remark is having some impact, he said.

A lot of “people already have their minds made up about what they want to do,” Rivara said. For others, however, the comment is “affecting what they thought.”

“You don’t joke like that,” he concluded.

A potential electoral college shakeup

Four years ago, Biden won the national popular vote by 4.5 percentage points, but won Wisconsin, which he needed for an electoral college victory, by just six-tenths of a point.

That gap has many analysts confidently saying that Harris needs to win the popular vote by around 3 percentage points to overcome the Republican advantage in the electoral college.

That GOP advantage, however, isn’t some law of political physics, and there’s a lot of reason to think the gap has shrunk — and maybe even reversed in Democrats’ favor.

The skew in the electoral college in 2016 and 2020 stemmed from the very large majorities Democrats won in the bluest states, including New York and California.

Their lopsided victories in such states gave Democrats tens of millions of extra votes that padded their nationwide popular vote but did nothing for them in the electoral college — a candidate doesn’t get any extra electoral votes for winning states by large margins.

Trump, by contrast, won his big states by relatively small margins.

This time around, however, the outsized Democratic majorities in big blue states seem to be shrinking, and Trump has been picking up more support in places where he already had a majority, such as Florida.

A NYT/Siena poll, for example, found Harris leading with New York City voters by 66% to 27%. That’s a lot, but it’s a good deal smaller than the 76% to 23% that Biden won in 2020.

Similarly, a UC Berkeley/Los Angeles Times poll early last month found Harris winning 57% of California’s vote, down from the 63% Biden took.

If those patterns hold, there won’t be much difference between the percentage Harris wins nationwide and her share in the swing states. So even a small Harris majority in the national vote could be enough to carry her to victory.

There’s even a small possibility that she could lose the popular vote and still carry the electoral college — a scenario that would cause whiplash across the political spectrum.

American politics has been stuck in the same bitter, close divide for more than a decade, and it’s tempting for people to think that the 2024 election will mostly recap 2020, although perhaps with a different conclusion.

Beneath the surface, however, a lot of change is going on. Sometime next week (probably), we’ll know how it all plays out.

What else you should be reading

Poll of the week: Americans’ top sources of political news ahead of the 2024 election

L.A. Times special: Here are the 100 California residents giving the most in the race for the White House

Was this newsletter forwarded to you? Sign up here to get it in your inbox.