The United States is set to hold its presidential election on November 5, and polls indicate that both major party candidates will enter Tuesday with a roughly equal chance at victory. The outcome will largely depend on turnout among key demographic groups and the results in several key swing states. While current polling in these states shows a slight advantage for Trump, the margin of error and complexities of the Electoral College leave Kamala Harris with a strong shot at overall victory. The Insider analyzes the balance of power and outlines the most likely scenarios.
While Election Day in the U.S. has been set for Nov. 5, most states have already begun early voting. As of the publication of this report, more than 64 million people have already cast a ballot (approximately one-third of the expected turnout). This is a significant number, and in some swing states, like Nevada and Georgia, turnout records for early voting have even been broken. The last election also saw extensive early voting (with a total of 100 million votes, of which 65 million were mail-in and 35 million in-person), largely due to the pandemic, as people avoided crowded polling stations. Democrats traditionally took the COVID-19 threat more seriously, while Republicans encouraged voting on Election Day, warning of the potential for fraud. As a result, mail-in ballots skewed heavily Democratic (estimated at 1.5 to 2 times more than Republican mail-in votes). Because state regulations required early ballots to be counted last, the initial Republican lead from in-person Election Day voting gradually narrowed, ultimately resulting in Joe Biden winning most of the swing states. For Trump and his supporters, this outcome reinforced their claims of widespread irregularities, leading Trump to reject the election results.
This year, there is no such early voting gap. In some swing states, such as Arizona, Nevada, and North Carolina, more registered Republicans than Democrats are even taking part in early voting. Additionally, a significant portion of votes will come from formally unaffiliated voters — while around 20% of voters fell into this category in the early 2000s, today, they make up more than 27%. Most of them, however, still vote for one of the two major parties rather than opting for a third party option.
Thus, predicting the outcome based on early voting by party affiliation is impossible, but it is safe to say that there won’t be as stark a contrast between early and late vote counts as there was in the previous election.
Victory hinges not only on attracting swing voters but also on maximizing turnout among one’s core supporters. The highest level of mobilization in the modern era was seen in the 2020 presidential election, with 158.4 million people voting — or 62.8% of the electorate.
Traditionally, it was believed that high turnout favored the Democrats more than the Republicans. Less politicized Americans tended to agree with Democrats, particularly on economic issues. However, in recent decades, an “education gap” has also emerged: Americans with higher education, who vote more frequently, now lean more Democratic, while less active, non-college-educated voters, especially among the white working class, have moved toward the Republicans.
As a result, high turnout now provides little advantage to the Democrats. According to a recent study, a 15% increase in turnout gave the Democrats a 1.5% vote boost in 2010, but only 0.4% in 2020. This shift relates directly to low-activity voter identification: in 2010, such voters included 7.2% more Democrats than the national average, while by 2020, there were 0.2% fewer. Numerous special congressional elections in 2023 appeared to validate this theory: in low-turnout elections, Democrats performed better than expected by rallying their most active voters.
This trend is especially significant in the context of a presidential election, where turnout is generally higher than in midterms due to the participation of less informed and more apolitical Americans. Some polls have shown that the Democrats’ strongest support is among those who voted in the last midterm elections and party primaries, while Republicans’ backing is strongest among those who voted only in the last presidential election — or skipped both.
Beyond turnout, Harris and Trump need to retain their electoral coalitions while also appealing to groups that previously voted for their opponents. In recent decades, Democrats have relied on support from minorities, primarily African Americans and Latinos, whose votes are crucial as their share of the electorate continues to grow.
However, in the last two presidential elections, this trend has weakened. For Democrats, polls taken when Joe Biden was still the candidate were especially concerning. While Biden won 92% of African-American and 61% of Latino votes in 2020, by the summer of 2024, his support among these groups had dropped to 77% and 54%, respectively — the worst results for a Democratic presidential candidate in decades. After Biden was replaced by Kamala Harris, who is of both Afro-Jamaican and Indian descent, Democratic support among racial minorities improved, though it still fell short of the levels seen in 2020 (84% among African Americans and 56% among Latinos). Fortunately for Harris, she has also gained ground among white voters, and since these still form the majority of the electorate, a two to three percentage point increase among them could offset some losses among minorities.
Although Trump has managed to sway some minority voters, a recent misstep may cost him the presidency. On Sunday, Oct. 27, at a rally at Madison Square Garden (MSG), comedian Tony Hinchcliffe referred to Puerto Rico as a “floating island of garbage.” Unsurprisingly, Puerto Ricans found the comparison offensive, and Pennsylvania (a key swing state) alone has more than half a million Puerto Rican residents. Prominent Puerto Rican public figures, including Jennifer Lopez and Ricky Martin, voiced their outrage, urging voters to support Kamala Harris. Trump’s response to the blunder was also questionable: speaking from a garbage truck, he claimed he neither knew nor recognized the comedian.
Trump’s weakest demographic is women — not only because his opponent is a woman, but also due to his many sexual scandals (from claiming that fame allows him to “grab women by the p*ssy” to cheating on his wife, who had just given birth, with porn star Stormy Daniels). One of the main issues in this election is abortion rights, which are no longer guaranteed nationwide due to a 2022 U.S. Supreme Court decision that overturned Roe v. Wade.
Democrats are pushing to restore universal abortion rights, while Republicans prefer leaving the matter to individual states. Beyond abortion, Republican messaging has a distinctly patriarchal tone: at a recent rally, former Fox News anchor Tucker Carlson earned the audience’s applause by portraying Trump as a stern father returning to give his daughter a “vigorous spanking” — an explicit reference to Kamala Harris. This sort of imagery appeals to older rural men, the core of the Republican base — yet it distances younger, educated women.
In the 2008 election, Democratic candidate Barack Obama secured 56% of women’s votes and 49% of men’s. Twelve years later, in 2020, 57% of women and 45% of men voted for Joe Biden. Now polls indicate the gender gap could widen to as much as 16%, and even higher in certain states. For example, a poll in North Carolina shows 61% of women and only 36% of men intend to support Kamala Harris.
Age is another important factor. Since the early 2000s, Democrats have received between 60-70% of young people’s votes (ages 18–29). Joe Biden, not particularly popular among younger voters, could have disrupted the trend, but Kamala Harris is expected to preserve it, aiming for support from around 60% of young voters. The issue is not that Trump, if elected, would turn 80 while in office — after all, Bernie Sanders is older than Trump but remains popular among younger demographics. Instead, the issue is that the agenda of Trump and his Republican party generally appeals to older Americans and is filled with references to how things in America were better in the past — from the strength of water pressure in the country’s toilets to its remembered might in international relations. This nostalgia, of course, is central to Trump’s best-known slogan, “Make America Great Again.”
When the factors of age and gender are combined, the gap becomes striking: Harris enjoys support from 67% of young women, and among college students, this figure exceeds 80%. Education itself, incidentally, is one of the key predictors in this election.
Gender and age are the most prominent dividing lines in this election, but according to some sociologists, the most reliable demographic predictor is education level. The higher the education level, the stronger the support for Harris: among Americans with a college degree, the Democratic candidate leads Trump by 21%, and among university professors, only 8% are willing to support Trump. In the scientific community, Trump and his policies are also unpopular, with 23 Nobel Prize-winning economists recently releasing a statement endorsing Harris’s platform as “vastly superior” to Trump’s, and an additional 80 Nobel laureates in various fields declaring that a Trump victory would endanger the future of science.
Of course, in electoral terms, Nobel laureates and university professors do not form a large enough voting bloc to directly influence the outcome, but their publicly expressed opinions impact segments of the electorate that regard them as authorities. And in key swing states like North Carolina, education also carries demographic weight, as the number of college-educated residents there has significantly increased in recent years.
Still, from an electoral standpoint, the primary opinion leaders are not Nobel laureates, but Hollywood stars and other entertainers. Here, Trump — himself a celebrity — might appear to hold an advantage. Yet on this front, he lags considerably behind Harris.
When assessing the electoral influence of the stars supporting Harris or Trump, it is essential to consider not only the celebrity’s prominence, but also the views of their audience. For example, Taylor Swift’s support for Harris is impactful not only because she is a global music icon, but also because her country-style songs resonate deeply in small-town America — a traditional Republican stronghold — while her home state of Pennsylvania is widely seen as the election’s most important battleground. Likewise, support from Detroit-born rapper Eminem holds weight, as he is a cultural icon in his home state of Michigan — another pivotal swing state. The ethnicity of these celebrities also plays a role; endorsements from Puerto Rican stars like Jennifer Lopez, Bad Bunny, and Ricky Martin may hold significant sway among a critical subset of voters.
Celebrities show support not only through statements but also through their performances: pop star Beyoncé took the stage at a Harris rally in her hometown of Houston, while rapper 50 Cent claimed he turned down a $3 million offer to perform at Trump’s recent MSG rally in New York.
A powerful (and potentially decisive) endorsement for Harris recently came from Arnold Schwarzenegger, who accused Trump of betraying fundamental American values by refusing to accept the results of the last election. “A candidate who won’t respect your vote unless it is for him, a candidate who will send his followers to storm the Capitol while he watches with a Diet Coke, a candidate who has shown no ability to work to pass any policy besides a tax cut that helped his donors and other rich people like me but helped no one else, a candidate who thinks Americans who disagree with him are the bigger enemies than China, Russia, or North Korea — that won’t solve our problems,” Schwarzenegger wrote in a post on X on Oct. 30.
The biggest issue for Trump is that Arnold Schwarzenegger is the only Republican who is more popular than he is in the U.S — and significantly so: 59% versus 42%, as per a recent YouGov poll. Republicans may easily brush off endorsements for Harris from liberal figures like Barack Obama or Taylor Swift, but Schwarzenegger — “the Terminator” and a former Republican governor — represents criticism “from within.”
For the same reason, criticism from John F. Kelly, a former Marine general who served as Trump’s Chief of Staff from 2017 to 2019, also proved particularly damaging. Gen. Kelly told a New York Times reporter that Trump “falls into the definition of a fascist” and added that Trump had privately commented — more than once — that “Hitler did some good things, too.” The latter remark quickly became one of the most discussed stories in the American media. Trump was also reminded of his past endorsement from a Ku Klux Klan newspaper and the rallies held by American fascists in the 1930s at New York’s Madison Square Garden — which Trump’s recent gathering resembled in both rhetoric and imagery. For example, Trump advisor Stephen Miller shouted, “America is for Americans and Americans only” at the Oct. 27 MSG rally — a near direct translation of the Nazi slogan “Only for Germans” (Nur für Deutsche), which was used to justify the Holocaust.
In response, late-night talk show host Stephen Colbert joked: “Now, to be fair, if I can make the argument, Hitler did one thing that was good: he killed Hitler. Hitler doesn’t get enough credit for that.”
Only a handful of notable celebrities have publicly supported Trump, and most of them aren’t A-listers. Hulk Hogan, the 1980s wrestling icon, made a show of tearing off his shirt at a rally in Milwaukee in July (though a similar attempt at the Oct. 27 MSG rally didn’t go as planned, as he struggled with the fabric). Actor and filmmaker Mel Gibson also endorsed Trump, claiming that Kamala Harris has “the IQ of a fence post” — a statement that raised few eyebrows given Gibson’s past controversies, which include blaming Jews for “all the wars in the world.” Among musicians, Trump’s most high-profile supporter is Kid Rock, who is arguably better known for his brief marriage to Pamela Anderson in 2006-2007 than for his music career.
The sole indisputable superstar at Trump’s rallies was Elon Musk, who not only spoke at one event but also offered financial backing — a significant gesture, given that Musk is the second-richest person in the world.
But Musk’s effect on Trump’s ratings remains unclear. Media attention centered less on Musk’s endorsement of Trump and more on his active promotion of discredited pro-Trump conspiracy theories on Twitter (now X) — including claims of fraud in the 2020 election.
Polls are plentiful, but they can be difficult for an inexperienced observer to interpret. National polls attempting to measure the popularity of Trump and Harris are the most straightforward, but given the importance of electoral distribution across states, they are not necessarily indicative of the outcome (Democrats have lost twice in recent history despite winning the popular vote). Still, if we take national polls as a general barometer and check aggregate data from sources like RealClearPolitics or FiveThirtyEight, we see that before Joe Biden withdrew, Trump held a solid lead, which narrowed sharply when Harris replaced Biden — though their ratings had leveled out again by October.
It’s worth noting that not all polls track only those voters who are certain to cast ballots. However, in this case, it doesn’t make much difference, as surveys of likely voters show similar results to those that gather a less targeted sample. More significant are the differences in methodologies and margins of error, which can skew aggregate data. FiveThirtyEight aims to account for these factors, but with Harris currently leading Trump by just over one percent overall, no methodology can be entirely reliable.
Additionally, it’s important to consider how many voters support alternatives to the Democratic and Republican candidates. Generally, the more support there is for third-party candidates, the worse Democrats fare. In his last two campaigns, Trump garnered no more than 47% of the vote. He won in 2016 largely because third-party candidates received nearly 6% of the total but lost in 2020 when their share fell below 3%.
Current polls show support for the Libertarian and Green Party candidates at a combined total of around 3-4%. In 2020, Biden received 51% of the popular vote, three points more than Hillary Clinton in 2016. This four-point lead over Trump was enough to secure Biden a narrow victory in the Electoral College. As Amy Walter, editor of The Cook Political Report, notes, this time around Trump is again expected to receive somewhere between 46-48% of the overall vote, provided that polling has been reasonably accurate.
In short, neither candidate enjoys a comfortable lead. Serious election projections are now focused on the complex landscape unfolding in seven key swing states.
The landscape of swing states is ever-evolving, with new states joining the list over time. North Carolina, historically a Republican stronghold, has recently shifted toward swing-state status due to demographic changes. This election's battleground grouping includes Arizona, Nevada, Wisconsin, Michigan, Pennsylvania, North Carolina, and Georgia. Currently, Trump leads in Arizona and Georgia by over 2%, suggesting these states will likely go Republican.
If this is indeed the case, it would lead to the following Electoral College map, in which Trump would need 24 Electoral votes from the gray states in order to claim the presidency:
The Democrats have a strong chance in Wisconsin and Michigan, while Republicans are almost certain to win North Carolina, bringing Trump’s total to 262. In this case, the outcome in Nevada is irrelevant, with Pennsylvania becoming the deciding state.
In Pennsylvania, Trump leads by half a percentage point, and the trend favors him — although Harris led just two weeks ago. However, this gap falls within the margin of error. Polls conducted by CNN and CBS in recent days indicate a dead heat.
In such a close race, any unexpected factor could tip the scales. For instance, offended Puerto Rican voters could swing in favor of Democrats, while Republicans may draw support from the Amish — a religious minority that usually abstains from voting. Although there are only about 100,000 Amish in Pennsylvania, if Republicans succeed in mobilizing a fraction of them, this could unexpectedly end up providing them the decisive margin.
Keep in mind that while the margin of error in federal-level polling is typically 3-4%, it is even higher at the state level. For the seven swing states, it averaged between 3% and 5.5% in recent elections. Pollsters have worked extensively to improve forecast accuracy after significant misses in recent contexts (such as overestimating Mitt Romney’s chances in 2012, predicting Hillary Clinton’s win in 2016, and forecasting a comfortable Biden lead in 2020 when the race was ultimately decided by just tens of thousands of votes). After the 2020 election, the American Association for Public Opinion Research concluded that they could not pinpoint the reason why polls had so significantly underestimated Trump. One theory suggested pollsters struggled to reach less politically engaged respondents, who tended to vote for Trump. This may explain why polls were relatively accurate in the 2018 and 2022 congressional elections, where turnout among such voters was typically lower.
Since then, the industry has made numerous adjustments. Many pollsters now use alternative data collection methods, such as text-message or mail-in surveys rather than phone calls, to increase response rates. New weighting methods have also been implemented to better account for underrepresented groups, and new companies employing alternative research methods have entered the market. Still, achieving a margin of error below 3% in state-level forecasts remains challenging.
This slim margin may be the only thing giving the Democrats’ grounds for cautious optimism.
Although Harris appears to trail in four of the seven swing states, with Nevada looking dead even, both candidates are well within the margin of error in each case. If each of these contests were to be viewed as a coin flip, the Democrats’ odds appear favorable, with 10 possible combinations bringing Harris to the necessary 270 Electoral College votes by winning just three of the seven. Trump, on the other hand, would need to win in Pennsylvania, Georgia, and North Carolina to take the election with only three swing states.
Thus, if swing state outcomes within the margin of error were randomly determined, Democrats would actually have a slight edge.
As Americans continuously refresh polling result pages, anxiety over the election runs high. A recent study by the American Psychological Association reveals that 69% of Americans report significant stress related to the election — a number that far surpasses the support either candidate can even dream of garnering on Tuesday.