Election polling experts in Eagleton webinar analyze current trends

On Tuesday, the Eagleton Institute of Politics held a webinar panel titled "Polling 2024: Expert Analysis on Election Polling and the State of the Race," moderated by Ashley Koning, the director of Eagleton's Center for Public Interest Polling.

The webinar included a panel of experts who discussed current polling trends and provided analysis and insight into these trends.

Don Levy, the director of Siena College Research Institute, who also works on the New York Times Siena Poll, said that the voters who will likely decide the election in seven key battleground states are mostly younger. These young voters are a focus group for pollsters, and they primarily care about the economy and reproductive rights.

"We certainly are reaching out to young voters very aggressively," said Levy.

The New York Times Siena Poll has also found that Democratic Senate candidates are performing better than Vice President Kamala Harris in battleground states while Republican Senate candidates are performing worse than former President Donald J. Trump.

"Now we're seeing some leads for Harris in the Rust Belt, Trump with some leads in the Sun Belt," said Levy.

Mark Hugo Lopez, the director of Race and Ethnicity Research at the Pew Research Center, spoke about crosstabs, the analysis of demographic groups and their effect on voting behavior and stances, specifically among Latino and Black populations.

Lopez said approximately 39 percent of Latinos support Trump this election cycle, a number that is higher than Trump's two previous elections.

"There's been a lot made about Hispanic men leaning towards Trump, and while it is a case that Hispanic men are more likely than Hispanic women to support Trump, it's really not that big a difference," said Lopez.

Lopez said that there will be a record turnout rate of Latinos because there are more Latinos who can vote this election year, although the effect of this on the election outcome is unknown.

Support for Trump is relatively low among Black voters at only 13 percent, Lopez said. 

Natalie Jackson, vice president of Greenberg Quinlan Rosner Research, a private polling company, discussed aggregates and forecasts in a pre-recorded video.

About polling aggregates, Jackson said, "It's still a nice tool. There are probably too many of them out there now. They all sort of say the same thing. You do have to be careful because they make a lot of assumptions."

Leading up to the election, the panelists agree that aggregates will be a useful tool in gathering information about where the election stands in a broad view, but it is important to always consider the margins of error.

There is a margin of error of 2 or 3 percentage points in any poll, said Lopez.

"I'd encourage any reader of polls to always look for the margin of error, and to be sure that they're thinking about it as percentage points and not necessarily percents," said Lopez.

Levy said that pollsters have faced challenges in recent years, especially regarding inaccurate pre-election polls in 2016 and 2020. Many voters no longer trust polling organizations.

"When Donald Trump is on the ballot, it has been a tougher terrain for pre-election pollsters," said Levy.

The New York Times is working to rectify this by asking about top-ticket preferences at the beginning of their survey and dealing with sampling errors differently.

Levy detailed the process that the New York Times Siena Poll uses to compile their data. Workers call registered voters in certain areas and must meet quotas for regions and demographics. They also often call people back multiple times in order to include them in their sample.

Joe Lenski, co-founder and executive vice president of Exit Polls at Edison Research, said that telephone polling has a 1 to 1.5 percent response rate, and email and other online polling forms have a 2 percent response rate. In contrast, in-person exit polling has a response rate of more than 40 percent.

Lenski detailed the exit poll process. These interviews are conducted in-person, over the phone and by email in the weeks leading up to the election for early and mail-in voters, as well as on Election Day at polling locations.

"Using these multiple modes are necessary to capture the full electorate and make sure there aren't any partisan biases in who we're talking to," Lenski said.

Jackson highlighted the differences between public and private polls. While public polling is meant to inform, private polling is meant to help in making decisions. She said that message development becomes important with private polling.

"You've got to, you know, step back sometimes, and say, is the conclusion that someone is drawing from a certain poll justified? So it's going to take a little time on the part of a consumer to fully appreciate the insights that are available from polls," Levy said.