Trump “beat” US opinion polls twice

As the United States awaits the results of the vote count from the remaining states on the battlefield, no matter who ultimately wins, the US polling industry has clearly repeated the mistake of four years ago of underestimating support. for President Donald Trump. This raises the question of whether the industry can survive another crisis of confidence.

“I want to see all the final results, I want to see the discrepancies between the pre-election polls and the final results,” said Christopher Borick of Muhlenberg University in Pennsylvania. “But there is a lot of evidence that we are in big trouble again. We need to see how deep they are.”

In some states, where pre-election polls predicted President Trump would lose by as little as Ohio, Iowa and Florida, he was practically the winner by a significant margin on the night of November 3. And in Biden’s state, which is supposed to win just as easily as Nevada, the gap is still too small to be determined a winner.

Based on the vote count, it’s clear polls overestimated the level of voter support for Bidenespecially among white and male voters. While pre-election polls say white voters over 65 have given up on Trump, that’s not really happening.

As a result, Biden did not perform well not only in various states like Florida, but also in white-populated suburban areas like Michigan’s Macomb County, where he was expected to outnumber him. principal.

Borick notes that while state polls were heavily skewed in 2016, they held steady during the 2018 midterm parliamentary elections. This led him to conclude that American views on Trump can be especially difficult to measure.

“Ultimately, like so many things to do with Trump, it may be necessary to run a different poll each time his name appears on the ballot,” Borick said. “Both elections with Trump are unusual compared to other elections.”

Comparing post-election polls to pre-election polls may provide some clues as to what hasn’t been done in pre-election polls.

Most surprising was the support for Trump from college-educated white voters, especially men. According to post-election polls, the two candidates received equal support from this group of voters, while almost all polls nationwide and in the state of the battlefield showed that Biden was dominant. Some studies show that highly educated people may hesitate to openly support Trump due to social pressure.

With respect to white groups without college degrees, in many pre-election telephone polls, Trump received mostly support from voters in their fifties. However, post-election polls show that voters in their sixties are the strongest support group, such as 2016.

What role does the electoral college play in American elections?

About COVID-19, It’s worth noting that, compared to most pre-election polls, post-election polls show that a smaller proportion of voters agree with conservative measures rather than quickly opening up society. As of the afternoon of Nov. 4, the number of voters who said blocking Covid-19 was more important was only 9% higher than those who wanted to quickly rebuild the economy, according to the post-election poll. In pre-election polls, this gap is typically in the double digits, and the majority of voters nationwide say they prefer to be cautious and control the epidemic.

It also appears that the virus does not motivate voters as much to choose candidates as many polls have broadcast. This year’s election has two data sources for the media to consider. The traditional dataset is a post-election poll conducted by Edison Research on behalf of a group of news agencies, conducted over the phone with voters who voted early and interviewed face-to-face at the polling place. Meanwhile, the AP conducts its own voter poll called VoteCast, which uses data from online responses collected by NORC, a research group based at the University of Chicago.

By looking at the two data sets, it can be seen that more voters regard Covid-19 as a major problem in their lives than those who say it is a matter of their own vote.

VoteCast showed that for every 10 voters, at least 4 consider the pandemic to be the most serious problem facing the country out of a list of 9 options. But in an Edison Research poll, when asked which issue has the biggest impact on the ballot, the number of respondents like Covid-19 is less than half the previous rate. The economy has a much bigger impact on them.

Not all pollsters have given bad results. Ann Selzer, long considered one of the top US polls, published data with the Des Moines Register a few days before Election Day, showing Trump a 7-point lead in Iowa, nearly identical to the results. reality.