US NEWS
Which political polls called the 2020 election correctly? What those polls say about the 2024 race
Election Day is just around the corner and everyone is looking for polls they can trust. A look at the agencies that got it right in 2020.
As the election approaches, many turn to the polls to determine where the race stands. However, before driving yourself crazy by looking at cross-tables, it is critical to remember that polls are snapshots. Some polled voters have every intention to go out and then don’t, while some who may report to a pollster that they are sitting out of the election end up casting a ballot.
The polls show a close race between Democrat Kamala Harris and Republican Donald Trump. Both candidates have an interest in the polls remaining close as they use those numbers to mobilize their respective bases. If the polls showed a knockout in one direction or another, the side projected to lose could see their votes fall off even more if they felt it was a lost cause. Similarly, the campaign trouncing in the polls could see voters stay home, thinking a victory is secure without them casting a ballot.
Don’t get caught up...
It is undeniable that Kamala Harris is in a weaker position relative to where then-candidate Joe Biden was in 2020. With Donald Trump in the White House and his handling of the pandemic coming into clear focus as he contracted the virus in the weeks before the election, the electorate began to show stronger signs that the Democrats stood a better chance of victory. However, when we look back at some of the polls, that strength was sometimes overestimated. For example, On October 7, Quinnipiac University published a poll based on surveys conducted between the first and fifth of October 2020 that showed Biden up by 11 points in Florida. Biden lost Florida by just over 37,000 votes or 3.36 percent.
This case highlights how race can change rapidly and the cyclical relationship between polling and campaign actions.
Upon seeing that poll, the Trump team may have poured more resources into the Sunshine State, and those efforts could swing many voters to their side. Less than a month later, another poll showed a much closer race, with Biden’s lead shrinking to 45 percent to 42 percent.
Focusing on the swing states
Since the Electoral College decides the presidency in the United States, this analysis focuses on the pollsters who correctly called the results in these six swing states: Arizona, Georgia, Michigan, North Carolina, Pennsylvania, and Wisconsin.
Keeping the 2020 map the same and removing the Electoral College votes for those six states puts Democrats at 232 votes and Republicans at 219.
CBS News and YouGov in Arizona
In 2020, the YouGov poll conducted with CBS and released on October 19 found that Biden held a three-point lead over Donald Trump in Arizona (50 percent to 47 percent). President Biden would go on to win the state by .30 percent, a much smaller lead than the poll suggested. Additionally, the poll’s margin of error was 4.6 percent, meaning there was still a chance that Trump could have won the state, and the result he received showed how close his campaign was.
YouGov and CBS released their most recent Arizona poll on October 18, showing Donald Trump with a slight lead over Kamala Harris (51 percent to 48 percent). The margin of error on that poll is ±3.3 percent.
If Donald Trump wins in Arizona, the party’s electoral vote jumps to 230.
Quinnipiac in Georgia
In mid-October 2020, Quinnipiac University’s pollsters showed Joe Biden ahead of Donald Trump by a 51 to 44 percent margin. He won the Peach State by 0.23 percent.
Quinnipiac‘s most recent poll, released on October 16, 2024, showed Trump ahead of Harris by seven points, almost the opposite result the pollsters reported in 2020. The margin of error is ± 3.0 percent, meaning Trump, according to this poll, holds a statistically significant lead over Harris.
If Quinnipiac calls the race correctly again this year in Georgia, the GOP’s electoral college count would increase to 246, with the Democrats holding onto 232.
Though concerning for Democrats, other polls that surveyed voters after the Quinnipiac results were released show a much tighter race. For instance, Atlas Intel shows Trump leading by one vote, and TIPP Insights shows the same result for Harris.
Morning Consult in Michigan
Many pollsters, including Morning Consult, correctly predicted Biden’s victory in the 2020 Michigan race. However, comparing the 2020 results to 2024 polling is problematic because many organizations that conducted polls four years ago have yet to release one for this race. Morning Consult showed Biden leading by seven points in late October, but he ended up winning by less than three points.
Currently, Morning Consult shows Harris with a two-point lead in the state, which is well within the margin of error.
If Democrats can hold onto Michigan, their electoral vote count will rise to 247, a virtual tie with Georgia.
Atlas Intel in North Carolina
Nearly all major pollsters got North Carolina wrong in 2020, but with Trump winning by 1.34 percent, it was anyone’s race.
In late October 2020, Atlas Intel projected Donald Trump to win with a two-point lead over his rival. Currently, the same agency has Harris winning by two points, based on surveys conducted between October 12 and 17. This polling group was one of the most accurate in 2020.
If Democrats flip the state, their electoral vote count will increase to 263, just seven votes short of the 270 needed to win.
Quinnipiac in Pennsylvania
In late October 2020, Quinnipiac projected Joe Biden to win Pennsylvania by a 51 to 44 percent margin. In November, he would win, but only by 1.17 percent.
On October 9, 2024, the organization released a poll showing Harris in the lead (49 to 47 percent) over Trump. This result is within the margin of error. Atlas Intel, who got the result incorrect in PA, is currently showing Trump with a three-point lead, and those surveys were conducted after Quinnipiac’s. Additionally, Atlas puts its margin of error at 2 percent.
If Democrats can secure Pennsylvania and the states mentioned above, they will win the election, surpassing the 270-vote threshold. However, if Trump wins, the GOP vote count will rise to 265, with all eyes on Wisconsin.
Atlas Intel in Wisconsin
The group projected correctly that Biden would win Wisconsin in 2020, and their latest poll shows Harris ahead in the state by less than a percent, with the margin of error for the survey coming in at ± 3 percent. By and large, the polls were much better for Biden in 2020 than for Harris today.
Based on the results above, whatever party wins, Wisconsin wins the election.