Compare the final NYT-Siena poll with the 2016 and 2020 elections

Final New York Times/A Siena College poll for the 2024 presidential election released Sunday shows Vice President Kamala Harris’ narrow lead over former President Donald Trump in all but one swing state.

The poll surveyed 7,879 likely voters by phone in seven battleground states. Among those polled, Harris led by three points in Nevada, by two points in North Carolina and Wisconsin, by one point in Georgia and had a virtual tie advantage in Pennsylvania, the poll found. Trump had a virtual tie advantage in Michigan and a four-point advantage in Arizona. Neither candidate had more than 50 percent support in any state, continuing to add to the appearance of a tight race with no clear advantage for either candidate.

The poll was conducted from October 24 to November 2, with a margin of error of plus or minus 3.5 points within each state.

However, Harris would have more positives than Trump. T survey analysishe’s the New York Times found that late deciders favored Harris, with 55 percent of those who had only recently decided their vote backing her, compared with 44 percent for Trump.

New York Times/ Siena College is one of the most reliable polling centers in America. Aggregator 538 ranks it number one on its list of 282 companies for its historical track record and transparency. Analyst Nate Silver ranks it among the top two firms, giving it an A+ grade.

Newsweek reached out to the Trump and Harris campaigns via email for comment.

What did the 2016 and 2020 surveys show?

The 2016 election damaged public opinion polls overall, as the vast majority showed a clear trend toward a significant lead for former US Secretary of State Hillary Clinton.

Many blamed then-FBI Director James Comey and his announcement just days before the election to reopen the investigation into Clinton’s emails for erasing what was otherwise a healthy lead and pushing the race within the margin of error.

On the eve of election day Time tracked Clinton with 45.9 percent support compared to Trump’s 42.8 percent support nationally.

Meanwhile, a national Siena College poll of battleground states showed Clinton had a seven-point lead in Pennsylvania, tied in North Carolina and even in Florida.

Pennsylvania famously carried Trump by 0.72 points (roughly just 44,000 votes), as did North Carolina by 3.66 points and Florida by 1.2 points.

The 2016 misfire led many pollsters to adjust how they weight different demographic groups to better capture the “Trump effect” in polls. However, these changes were not as effective as sociologists had hoped.

The 2020 Siena College poll found Joe Biden with a national lead over Trump among likely voters by nine points. Biden won by 4.5 points, only half of what the polls showed.

In the 2020 public opinion poll released on November 1 of that year, Siena College found Biden ahead by six points in Arizona, three points in Florida, six points in Pennsylvania, 11 points in Wisconsin.

Biden ultimately won Arizona despite broader expectations of a Trump win, but only achieved a stealth victory by 0.3 points — roughly 10,500 votes. Trump easily won Florida by 3.5 points, and Biden took back Pennsylvania, but again only by a narrow margin of about 1.16 points.

Biden also ultimately took Wisconsin, but by a much smaller margin of 0.63 points.

Can we trust polls?

The last two elections have seen polling suffer serious reputational damage, with polls showing Clinton would beat Trump in 2016 and that Biden had a larger lead than he won in 2020.

“Although the polls are not so accurate that one can be confident that they will have a difficult election, it cannot be assumed that the polls will again be wildly off, as they were in 2016 or 2020,” Cohn, chief political analyst. For Time– wrote on Friday in an article published in Upshot.

Both previous polls underestimated support for Trump, and in 2016, many found that his supporters were less likely to express their support for the highly controversial candidate.

Cohn highlighted “major methodological changes” in the polling data, noting that “many of the top-performing polls of 2020 either adopted massive methodological changes or disappeared from the map.”

However, some of those changes, made to compensate for Trump’s underrepresentation, could instead result in Harris being underrepresented, Cohn warned.

“It is difficult to overstate how traumatic the 2016 and 2020 elections were for many pollsters,” he wrote. “For some, another underestimation of Trump could pose a serious threat to their business and livelihood. For others, their status and reputation are at stake.”

He added: “If they underestimate Mr. Trump for the third time in a row, how can you trust their polls again?” he continued. “It is much safer, whether from a literal self-interest standpoint or purely psychologically, to find a close race than to bet on a clear Harris victory.”