In the weeks and days leading up to the 2024 presidential election, polls generally reflected a very close race, with voters divided more or less evenly between donald trump and Kamala Harriseven in the seven disputed states.

There even seemed to be a push in favor of Democratic candidate Harris toward the end, with some of the country’s top forecasters seeing enough in the polling data to place her as the marginally most likely winner in their election simulations.

One particularly shocking poll by Ann Selzer in Iowa, a highly-rated pollster with a strong track record, showed a big swing toward Harris in the final week of the race, putting her three points ahead of Trump in a state she won. by nine in 2020.

In the end, Trump won all the swing states and is on track to win the popular vote-the first Republican the candidate will do so in 20 years, if so, and may have a Republican majority Senate and House too. Some votes are still being counted, including millions in California.

He also won Iowa by 13 points.

There is a sense in the post-election autopsy – fueled in part by how forecasters used the polls – that this is another mistake by the polling industry because it generally failed to capture the strength and breadth of support for Trump across demographic sectors. clue.

But is this true? Were the polls wrong in 2024? If so, why? Or is this a mistaken assumption and the polls were much more accurate than they are given credit for? news week pollsters and public opinion experts asked. This is what they said.

Patrick Murray, Director, Monmouth University Polling Institute

The polls (if you really paid attention to them and not to the forecasters who misuse them as crystal balls with pinpoint accuracy) told the story of a stable and uncertain race that could easily swing in either direction by a few points. That’s exactly what happened.

John Zogby, Senior Partner, John Zogby Strategies

Objectively speaking, no one really “got it” if you look for that precise number. Or, in any case, even catch the trend. If I had to give an award, I think TIPP had Trump ahead by one. For our part, in the two-way race we had Harris up by two.

But I know from my surveys, and the few I looked at by others, that we captured the demographic trend line. We all saw that there was going to be a substantial gender gap, and an even bigger gap between young women and men.

For the most part we saw the Latino vote change. In fact, we had Trump ahead among Latinos, in the mid-40s, a point or two ahead of Harris. And young black people: we have it.

In summary, I think the surveys were useful and should never be discarded. The theme of my book, Beyond the horse raceis that we should not hang our hats on the horse race, and the race for the horse race. But in reality, we must pay attention to what it tells us.

We need to focus much more on what surveys tell us about ourselves and others. Am I the majority? Am I in the minority? Am I completely alone? And what drives what I do?

Everyone saw that the country was going in the wrong direction and that a sitting president was stuck with, at best, a 40 percent job approval rating. But the depth and breadth of the Trump was quite surprising.

Chris Jackson Senior Vice President, US Public Affairs, Ipsos

I definitely disagree with the idea that the polls were wrong. Multiple polls were released showing Trump winning these states, and it was clear to any registered pollster that the election would come down to which side would turn out best for his base. The vote totals suggest that Trump was able to retain his 2020 voters and increase them by 1 to 2 percent, while Harris appeared to struggle to match Biden’s 2020 numbers.

Mark Penn, President and CEO of Stagwell

I think the polls were good overall, but not great in the sense that they told (with the exception of Ann Selzer) which groups were moving and what issues people were concerned about. Like last-minute predictive machines, they needed to adjust about 2 more points.

I think my assessment of the polls as in the Wall Street Journal It was that Trump had the advantage to win and he did it.

Christopher Wlezien, Hogg Professor of Government, University of Texas at Austin

I actually think the polls got pretty good results and the performance may improve as the vote counting continues.

Using the final 538 figures, when I took a look at the Nov. 6 votes, the national polls appeared to be off by 2.4 percentage points and the swing state polls by about 2.6 points on average.

These may not be the best numbers, but they are below the average errors, that is, as previously suggestedThe average error in last week’s national polls for all presidential elections between 1952 and 2020 is 2.5 points, and it is even higher in the states.

Polls still consistently underestimate Trump’s vote share, as we’ve seen before, and this led to incorrect “calls” in some states and probably the entire nation, depending on where we end up.

But it’s obviously difficult to nail down the winner when the races are so close, even when the polls are doing very well.

I’m interested to see what we can glean from the information about what polling organizations did to produce their estimates and whether and how this affected their results and performance, which might also help us understand the tendency to underestimate Trump’s share of the vote.

Mike Traugott, Research Professor Emeritus, Institute for Social Research, University of Michigan

By one measure of precision, national polls generally did well, meaning most showed a very close race and the result reflected that.

While the typical difference between candidates was around 1 or 2 percentage points, sometimes as much as 4, the actual difference can be 1.6 percentage points compared to the split of the popular vote.

On the other hand, there appears to be a continued underestimation of support for Donald Trump. It will take some time to determine to what extent this was due to unsatisfactory likely voter models, the adequacy of new weighting algorithms, or “shy Trump voters” who will not participate in polls or say they intend to vote. for him.

For the polling industry, it is important to determine whether this is endemic to Donald Trump as a candidate or reflects a more fundamental methodological issue.

Christopher Borick, director of the Institute of Public Opinion at Muhlenberg College

Actually, the polls as a whole performed quite well this cycle. If you look at the final national popular vote averages and swing states, the poll estimates were pretty close to the true mark.

There was once again some degree of underestimation of Trump and some other Republican candidates in the cycle, marking the third consecutive cycle of errors in the same direction.

Unlike 2022, where academic pollsters and mainstream media polls seem to have done better, this cycle it seems like some fairly new pollsters, like AtlasIntel, performed really well.

Josh Clinton, co-director of the Vanderbilt survey

Overall, the polls once again appeared to underestimate support for Trump across the board, just as they did in 2016 and 2020.

If the pollsters wanted to try to get a positive result out of the night, the amount they lost was less than in 2020; Based on preliminary results, it appears the polls underestimated Trump’s margin in swing states by about 2 to 3 percentage points.

But it doesn’t seem like that interpretation is going to make anyone really happy. Close, but without a cigarette.

I think this really emphasizes again the difficulty of pre-election polling: when you survey 800 people, the responses of 8 people can cause a 1 percent error.

And when it is necessary to call 40,000 people for 800 to take a survey (2 percent response rate), it becomes clear that differences in who responds or in what pollsters assume about the electorate can produce errors that turn out to have consequences when trying to predict a 1 percent poll. -2 point race.

In general, I think pollsters and the public expect too much from pre-election polls and that polls are better suited for problems where small errors have fewer consequences; It doesn’t matter much if support for a policy is 57 percent. or 65 percent since they are both “high,” but that’s important when trying to predict a close race!

Courtney Kennedy, Vice President of Methods and Innovation, Pew Research Center

People often overlook the things that polls got right. Polls clearly show a dissatisfied electorate. They showed that voters were deeply focused on the economic pain caused by inflation, even as pundits touted a variety of positive economic statistics, from a relatively low unemployment rate to gains in the stock markets. And polls showed that more voters trusted Trump than Kamala Harris to fix the economy.

Pre-election polls also showed widespread concerns among Trump supporters about the impact of illegal immigration. In fact, for Trump supporters, immigration was among the top issues of the election, according to our own pre-election polls.

And polls showed that a sitting president clearly falls short of public approval. No sitting president has won re-election with an approval rating as low as joe biden‘s.

It’s fair to say that the polls did not indicate the extent of Trump’s victory, although the possibility was there.

Polls in battleground states were generally correct that those seven races would be decided by a few percentage points. And it was always possible, however unlikely, that Trump could win most or all of them.

The polls weren’t perfect, but this wasn’t 2016. It was 100 percent clear that Trump could win. And for people who understand polls, it was possible he could sweep battleground states, some of which are still counting votes.