Every presidential election appears to pose one big question—who will win?—that is in fact made up of countless smaller questions: How do voters really behave? Which old rules of politics still apply, and which are obsolete? What kind of country do we live in? In 2016, we learned that white evangelical voters would overwhelmingly support a louche serial philanderer. Four years later, we learned that Florida had shifted from the quintessential swing state to a Republican stronghold. Here are five of the biggest outstanding questions heading into next week’s vote.

Will the polls finally be right?

Donald Trump’s stunning 2016 victory set off a reckoning among pollsters to figure out how they had gotten things so wrong. Then 2020 came around, and they somehow did even worse. Polling averages showed Joe Biden leading in Wisconsin, for example, by 10 points; he won the state by just half a point.

Pollsters have offered various overlapping explanations for their errors last time. Republicans seem to have been less likely to respond to surveys, because of a deep mistrust in institutions, which left them underrepresented in the results. And Democrats may have been more likely to respond, because they were more likely to be sheltering in place during COVID. Whatever the precise mechanism, the 2020 polls clearly underestimated support for Trump.  

[Gilad Edelman: The asterisk on Kamala Harris’s poll numbers]

In 2024, pollsters have been deploying a range of techniques to prevent that from happening again. One common approach: asking people whom they voted for in 2020 to ensure that surveys include enough Trump 2020 supporters. Such techniques, however, can introduce problems of their own. Voters are bad at recalling past votes, and tend to say that they voted for the winner of the previous election even if they didn’t. This raises the possibility, however remote, that polls are overestimating Trump’s support this time around.

Will we finally see a youth gender gap?

In an electorate deeply divided by race, class, geography, and education, gender has long been an exception. Since the 1980s, men have been slightly more likely to vote Republican and women to vote Democratic, but the gap has remained small and stable. Among young voters, it has hardly existed at all; young people have skewed overwhelmingly Democratic regardless of gender. In 2020, 68 percent of 18-to-29-year-old men voted for Joe Biden compared with 70 percent of women in that age cohort. That was the same percentage gap as in 2008.

If the polls are to be believed, that pattern has radically changed this year. Across three recent New York Times/Siena polls, young women still support Democrats at about the same rate as they did in 2020, with 67 percent in favor of Kamala Harris. But young male support for Democrats has plummeted to just 37 percent. In swing states, the gap appears to be even larger.

What makes this shift especially strange is that its sudden timing rules out many of the most common explanations offered for it. The backlash to #MeToo, Trump’s hypermasculine appeal, changing gender roles, and the rise of an anti-establishment male online subculture have been many years in the making, and yet the youth gender-voting divide didn’t show up in 2018, 2020, or 2022. Why it might be showing up now remains a mystery. (It also doesn’t seem to be about the gender of the Democratic candidate; Joe Biden was polling just as poorly with young men as Harris is.)

[Rose Horowitch: Are Gen Z men and women really drifting apart?]

The possibility remains that the divide is an artifact of polling that will not extend to the voting booth. Trump’s youth support is concentrated among those who are least likely to actually vote. According to the most recent Harvard Youth Poll, young men who “definitely” plan to vote favor Harris 55 to 38 percent. Young men might say they prefer Trump, but whether they will act on that preference is a different story.

Are Democrats losing Black and Hispanic support?

The American electorate has long been sharply divided on racial lines. Since the 1960s, white voters have mostly voted Republican and nonwhite voters have overwhelmingly voted Democrat. In 2020, Joe Biden won 92 percent of Black voters and 63 percent of Hispanic voters. Hillary Clinton and Barack Obama performed similarly among those groups.

Four years later, Trump’s rhetoric toward nonwhite Americans and immigrants has become even more nakedly hateful, while Democrats have nominated a Black woman for president. And yet, according to a recent New York Times/Siena poll of the Black and Hispanic electorate, Harris is winning just 78 percent of Black voters and 56 percent of Hispanic voters. If those numbers hold on Election Day, Trump is on track to win a greater share of Hispanic voters than any other Republican candidate in two decades and a greater share of Black voters than any other Republican candidate since the passage of the 1964 Civil Rights Act.

How can this be? One possibility is that economic concerns are overwhelming racial ones. Black and Hispanic voters have long been more likely than white voters to say the economy is their top issue, and right now the country’s economic mood is dismal. The same Times/Siena poll found that just 20 percent of Hispanic voters and 26 percent of Black voters say current economic conditions are good or excellent.

Another possibility is that the same forces that first caused white voters without a college degree to swing toward Trump in 2016 are now causing nonwhite voters to do the same. Many Black and Hispanic voters agree with Trump on issues such as immigration and crime: The Times/Siena poll found that 45 percent of Hispanic voters and 41 percent of Black voters support deporting undocumented immigrants, and about half the voters in each group say that crime in big cities is a major problem that has gotten out of control. And both groups have become disillusioned with the Democrats. The poll found that just 76 percent of Black voters and 56 percent of Hispanic voters see them as “the party of the working class,” while only six in 10 Black voters and fewer than half of Hispanics say that the Democratic Party “keeps its promises” more than Republicans.

[Read: How Trump is dividing minority voters]

Either way, a shift of this magnitude would overturn two interrelated assumptions that have dominated the thinking of both major parties for decades: first, that voters of color predictably vote according to their racial identities, and second, that as the U.S. continues to become a more racially diverse country, the electorate will automatically tilt in favor of the Democrats. A political system in which nonwhite voters are truly up for grabs has the potential to reshape the strategies of both parties and transform the electoral map.

Does the economy matter?

Historically, the state of the economy has been a pretty good predictor of who will win the presidency. An analysis by the political scientists John Sides, Chris Tausanovitch, and Lynn Vavreck found that despite all the abnormalities of 2020—a pandemic, national protests, a uniquely polarizing president—models that factored in both economic fundamentals and consumer sentiment predicted the result and margin of that year’s presidential election more accurately than the polls did.

That should be good news for Harris. By most objective standards, the U.S. economy is performing remarkably well: Growth is up, unemployment is low, real wages are rising, and inflation has been tamed.

Except the voters seem to disagree. Despite a stretch of fantastic economic news—including interest-rate cuts, low inflation, plunging gas prices, and continued job growth—consumer sentiment remains well below where it was as recently as April of this year and at about the same level as it was in October 2009, when the economy was in freefall and the unemployment rate reached more than 10 percent. Even as the economy has improved in almost every possible way, voters don’t seem any happier with it. Many Americans are still outraged by the higher cost of goods, particularly groceries, relative to pre-pandemic prices. And, like voters around the world, they seem likely to take that frustration out on the incumbent party.

[Annie Lowrey: The worst best economy ever]

But here’s a further twist: Polls also show Harris’s standing improving along the specific dimension of economic issues. Every month this election cycle, the polling firm Echelon Insights has asked voters which candidate would make the economy work better. In June, voters favored Trump over Biden by 11 points; in September, they favored Harris over Trump by one point. That might help explain why Harris is doing better in the polls than Biden did, but it doesn’t explain the fact that Trump has been gaining ground in recent weeks to pull dead even with the vice president, even in some national polls. The relationship between the economy and voting behavior in the 2024 election appears to be anything but straightforward.

Do campaigns make a difference?

The core of every campaign is what’s known as the “ground game”: each side’s effort to canvass neighborhoods, knock on doors, and make phone calls in an attempt to turn out its supporters come Election Day.

But the ground game has been a remarkably poor predictor of success in recent elections. In 2016, Trump’s field operation was almost nonexistent, whereas the Hillary Clinton campaign oversaw a voter-outreach juggernaut. Trump won. In 2020, the Trump campaign boasted that its massive field operation knocked on a million doors every week, while the Biden campaign conducted almost no in-person canvassing because of worries about spreading COVID. Biden won.

Still, political-science research has consistently found—and common sense strongly suggests—that nudging potential voters to vote does, in fact, increase turnout. According to estimates by the political scientists Alan Gerber and Donald Green, a canvassing effort that gets a response at 1,000 doors generates about 40 new voters, and a phone bank that reaches 1,000 people produces approximately 28 new voters. Given that the 2024 presidential race could very well be decided by tens of thousands of votes in a few key states, these kinds of numbers could be enough to swing the outcome.  

So who has the better ground game this time around? By just about every conventional indicator, the answer is Kamala Harris. The Trump campaign claims to have “hundreds of paid staff”; the Harris campaign has 375 in Pennsylvania alone, and about 2,500 in total. During just one week in October, the Harris campaign says its volunteers knocked on 1.6 million doors and made 20 million phone calls. (Trump’s team has chosen not to release these kinds of details.)  

The disparity is partly a product of an imbalance in resources. The Harris campaign has raised more than $1 billion in the past three months, more than double the Trump campaign’s haul during the same period. The Harris campaign accordingly outspent the Trump campaign by more than three to one in September alone. (Making matters worse for Trump, his campaign has spent a large chunk of its war chest paying off his legal bills and funding efforts to monitor “election integrity.”)

The Trump campaign says it can make up for its lackluster on-the-ground numbers by relying on unconventional tactics, such as hyper-targeting “low-propensity voters” who support Trump but didn’t show up in 2020. It is also relying heavily on well-resourced but unproven outside organizations funded by conservative donors to get out the vote.

Judging by the past two elections, odds are that Trump’s lack of a ground game won’t be decisive. But in an election in which almost every single swing-state vote might count, it certainly isn’t doing him any favors.