Decision 2020: Trying To Get Polls Right After 2016 Miss


Many polls underestimated President Donald Trump’s support in 2016, in part because some voters hid their preferences from those asking the questions at that time.

Four years later, are Trump supporters less bashful?

"Lots of people were really uncomfortable talking to people (in 2016)," said Jill Darling, Survey Director of the Daybreak Poll at the USC Dornsife Center for Economic and Social Research.

That’s one of the interesting nuggets in the latest USC Dornsife Daybreak Poll, an ongoing survey conducted of more than 5,000 voters who are taking part in USC’s Understanding America Study.

"What we found is that people who made up their minds to vote for Trump and people who were in rural or small towns were the least comfortable in discussing their vote preferences," Darling told KCBS Radio’s "The State Of California" podcast. "They were also tending to be the latest deciders on their votes."

Many of those people voted for Trump in 2016 than were expected to do so.

This year, the results may be different.

"We’re seeing a small amount of party switching among Republicans, not really enough to make a significant amount (of a difference)," Darling explained. "But what we are seeing is a lot of changes between how people are voting who are in the categories of, they say they’re independent, but then if you push them a little bit they’re leaning towards the Republicans or they’re leaning towards voting for Trump. Those are the groups that are the softest."

Much of that group was in the Trump camp four years ago.

Darling explained that Daybreak Poll data showed the same group leaning Democratic after the Democratic National Convention, but then swaying back to voting Republican following their convention just a week later.

Jobs and the economy remains the top issue in the campaign, according to Darling.

The poll showed Biden with a 14-point lead as recently as August 29.

That lead is now down to seven points in data released Tuesday.