As more results are finalized from Tuesday’s election, it looks like the polls once again underestimated Donald Trump’s support nationally and in key battlegrounds. But compared to the major misses in 2020 and 2016, pollsters made noteworthy improvements this year.
If one had only listened to the pundits in the weeks leading up to Election Day, there were two big “surprises” as the votes came in. The first was that Trump swept all seven swing states, handing him a landslide Electoral College victory. The second, and perhaps more unexpected result, was that Trump won the popular vote.
But neither of these results should’ve been all that surprising given the polling data. In the RealClearPolitics average, Harris had a national vote lead of just 0.1 points on Election Day eve. Most polls ranged from Harris +2 to Trump +3, indicating a close race that could go either way given that those polls had a margin of error between one and three percent.
While votes are still being counted, The New York Times projects that Trump will finish with a 1.5-point popular vote advantage. That result is well within the margin of error for most major polls. That’s a far cry from 2020, when the final average showed a 7.2-point advantage for Biden and he won a 4.5-point victory.
The fact that Trump won every swing state was another “surprise” that shouldn’t have been all that surprising. In the RCP averages, Trump led in every swing state except Michigan and Wisconsin, where Harris led by half a percent. Once again, a Trump victory was still within the margin of error in those states.
Nonetheless, the polls still underestimated Trump’s support in every swing state. In Pennsylvania, Trump’s actual margin was 1.5 points higher than his advantage in the last RCP average. In Georgia Trump did 0.9 points better, in North Carolina he did 2.2 points better, in Michigan he did 2 points better, in Wisconsin he did 1.3 points better, in Arizona he did 2.4 points better, and in Nevada he did 4.2 points better – by far the biggest polling miss in the seven swing states.
These elections were still close, although Trump won Wisconsin and Georgia by larger margins than Biden did in 2020. On average, the battleground polls underestimated Trump’s support by about 1.3 points – enough for him to carry every swing state. This was a marked improvement from four years ago, when state-level polls underestimated Trump by sometimes as much as six and seven percent.
What the polls appear to have missed was the breadth of Trump’s support not just in the battlegrounds, but nationwide. Trump winning any one swing state in itself is not surprising – the fact that he won all seven points to a deeper trend among the electorate that most pollsters did not predict.
Trump’s performance vs. the polls was notably better in the Sun Belt than the Upper Midwest, a trend to pay attention to in future cycles. Nevada saw a seven-point swing from the 2020 results there, with Trump running far ahead of the polls. In Texas, Trump won by 14 after polls showed him up just nine. It was the same story in Florida, with Trump winning by 13 after polls showed him up eight.
Ohio was another pleasant surprise for Trump that pointed to deeper support for Trump nationally than many polls indicated. After Trump won the state by about eight points in 2016 and 2020, most polls showed a similar result this year. The actual result was that Trump won by more than 11 points, continuing a rightward trend in what was once a perennial swing state.
In hindsight, Trump always should’ve been considered a strong favorite heading into Election Day. Harris’s sudden entry into the race in late July gave her a honeymoon boost in the national polls as the media heaped praise on her and a reinvigorated Democrat machine tried to boost her over the top.
But in the battlegrounds Harris struggled to gain an advantage over Trump, running neck and neck with him throughout August, September, and early October. By mid-October Trump had seized back the momentum and the initiative, closing the gap in the national polls and building a lead in the battleground polls.
There will certainly be plenty of well-founded discussions in Democrat circles in the weeks and months ahead about how so many believed so strongly that Harris had a clear path to victory, and even that she was a favorite in the race.
At least part of this belief appears to have been founded on biased outlier polls and a willful ignorance about what the reliable polls were actually saying. Having worked so hard to convince the public that Trump is the embodiment of human evil, Democrat insiders seem to have convinced themselves that no one could possibly vote for him.
This time, however, Democrats who feel blindsided by the result can’t blame woefully inaccurate polls as they did in 2016. A Trump victory was always staring them in the face – they just refused to see it.
Shane Harris is a writer and political consultant from Southwest Ohio. You can follow him on X @shaneharris513.
