SAN JOSE (KPIX 5) – The presidential election is still hanging in the balance as votes continue to be counted in states across the country, but polls leading into the race painted a very different picture.
Political experts say this year was a very different election year for a number of reasons. The pandemic lead to record levels of mail-in ballots and early voting, which could be part of the reason why the polls were so off. Those same experts say that doesn’t explain why the polls got it so wrong two presidential elections in a row.
A few weeks before Election Day 2020, dozens of polls showed Joe Biden as the clear leader over President Donald Trump in many battleground states. It also showed Democrats were likely to gain the majority in the Senate and pick up seats in the House.
Based on the votes that have been counted so far, those polls are off.
“Clearly, Trump’s support was underestimated again,” says Melinda Jackson, the Associate Dean of the Political Science Department at San Jose State University. She is an expert in political psychology and polling.
Jackson acknowledges the Presidential polling was off in 2016 and 2020 and says one of the biggest reasons why is because both years Trump was on the ballot.
“Donald Trump frequently talks about not trusting the media, when a lot of these polls are done by media organizations and universities, you could see how a Trump supporter might think, nope, I’m not participating in this,” says Jackson.
If that is in fact what happened, that means the polls would be off and skewed in favor of Joe Biden.
“These are not partisan polls. The purpose is to get it right. They don’t have an interest in getting it wrong,” says Jackson.
Another place the polls got it wrong this year was in House and Senate races, but Jackson says that is also because she thinks Trump supporters aren’t likely to talk to pollsters.
“They’re not just going to ask you one question when they call you up. They’re going to ask you about the Presidential and the Senate race in your state and any House races, so if the person who is supporting Trump is not participating in that poll, then all of their other vote preferences are not going to be captured either,” she says.
Jackson is quick to point out that after 2016 pollsters looked at where they thought they went wrong and made changes in their methodology for polling during the 2018 midterm elections.
The results of the 2018 elections showed the polls were right on with their predictions, so pollsters thought they’d fixed the problems.
Fast forward to the 2020 presidential election and the polls are off again, so those same pollsters will be going back again to try to figure out what happened and where they went so wrong.